Bug: http://b/77792074
- Add the libclang_rt.profile runtime libraries directly to the compile
command (for both host and target) instead of relying on the Clang
driver.
- Move the coverage mutator to PreDepsMutators so the mutation has
already happened when runtime libraries are added during dependence
computation.
- Factor out cc/config/toolchain to identify libclang_rt.profile modules
for the x86 and x86_64 host.
Test: make NATIVE_COVERAGE=true produces coverage-enabled host binaries.
Change-Id: I1ebc8cffdf11622bfc18199a57674672888b3a5f
Bug: http://b/73768157
Test: Mark 'use_clang_lld: true' for one arch of a module and verify by
inspection that only that arch variant uses lld.
Change-Id: I4b0d7a2bc43ee0b8462ab09a5f5c50807129a7fb
This change is to let people to just copy & paste the update command
Test: break the ABI of libbinder intentionally and make.
The message includes absolute path to the create_reference_dump.py tool.
Change-Id: If7703d64be1f51ea4b054aa1e815276e1d6dbc14
Bug: http://b/64121881
Bug: http://b/78188880
- Allow filegroup's properties to be extended by a LoadHook
- Support a filegroup (':module') in a prebuilt's 'Srcs' property to
export files from a different path as the prebuilt's sources.
This change also includes a refactoring that moves genrule/filegroup.go
to android/filegroup.go so that FileGroupFactory is visible in
prebuilt_test.go.
Test: Test
https://android-review.googlesource.com/c/platform/development/+/469159
in clang-tools branch on Linux, Darwin. Test regular build in
aosp/master.
Change-Id: I3ff6215ab2e62955f039fd1086c31f1bd50ebcf6
Set LOCAL_NATIVE_BENCHMARK to true. Then, build system can use it
to auto generate the related Test config file of Google Benchmark
Test
Bug:74844131
Test: Use mm to rebuild any module in cc_benchmark
Change-Id: I6a8f1ff78f7aec3f4eb17e218845f0cdf3f57f32
double_loadable is a property that tells whether a module is capable or
being loaded with other instance (possibly an older version) of the same
module in the same process. Currently, a shared library that is a member
of VNDK can be double loaded in a vendor process if the library is also
a dependency of an LLNDK library. Such libraries now must be explicitly
marked as `double_loadable: true` by the owner, or the dependency from
the LLNDK lib should be cut if the lib is not designed to be double
loaded.
Bug: 77155589
Test: m -j
Merged-In: I3b839f860cbdc01f43b59872cd7bb84ac4a7d73e
Change-Id: I3b839f860cbdc01f43b59872cd7bb84ac4a7d73e
(cherry picked from commit 89943d8be2)
* Add -fuse-ld=lld to hostGlobalLldflags.
* Set up *ClangLldflags variables for hosts.
They are the same as *ClangLdflags, but could be changed in the future.
Bug: 73768157
Test: make checkbuild
Change-Id: I3d61504ab7262d472bbf933df7c1a9bef62519e1
Bug: 78118272
Bug: 72225642
Test: SKIP_ABI_CHECKS=true mm -j64 in external/libjpeg-turbo; no abi
dumping / diffing happens
Test: SKIP_ABI_CHECKS=false mm -j64 in external/libjpeg-turbo; abi
dumping / diffing happens
Test: SKIP_ABI_CHECKS=foo mm -j64 in external/libjpeg-turbo; abi
dumping / diffing happens
Change-Id: I6330bc6de81abd589e78572af8efdf70d4c69b80
When we're doing ABI diffing, we run something like:
(run-abi-diff) || (echo && exit 1)
When `dist` is added, we attempt to copy out the result:
(run-abi-diff) || (echo) && (copy && exit 1)
But this always fails, since it will always run exit 1. We really want
the parenthesis around the entire second section. This change switches
it to:
(run-abi-diff) || (echo && (copy) && exit 1)
This matches the behavior before Iae25374fe937a0cbe8a8ddf9e23c3bc1f62bbb2a
Test: treehugger on P
Change-Id: Ic39da4d4630ee950b811ec7854d55d65c52f8661
* USE_CLANG_LLD is unedefined in current builds.
* When USE_CLANG_LLD is defined to 'true' or '1',
use clang's lld instead of ld or ld.gold.
* When lld is enabled:
* ld-only flags are not passed to 'lld'.
* location_packer is disabled.
* Use new lld's --pack-dyn-relocs=android.
* When lld does not work:
* In Android.mk files use LOCAL_USE_CLANG_LLD := false.
* In Android.bp files use use_clang_lld: false.
* Only arm, arm64, x86, and x86_64_devices have LLD flags;
all other hosts and targets do not call lld yet.
Bug: 73768157
Test: make checkbuild and boot
Change-Id: I06b8a1e868a600997a7e70fe05c299d751d23d5f
During link-type check, if a module with sdk_version "current" refers
another, compatibility check between STLs has not been performed.
Bug: 77834464
Test: m -j succeeded
Change-Id: Id82a39372670daca779d4fb4af2deb202170a2fd
In preparation for unexporting ProductVariables, explicitly return a
pointer to the structure from TestConfig / TestArchConfig.
Bug: 76168832
Test: m blueprint_tools
Change-Id: Iccfb4c912f8e0ee3f620cc1ee00f0cdc5cba7735
An upcoming change will stop exporting ProductVariables from Config, so
switch to using existing accessor functions, and add more when they're
missing.
Bug: 76168832
Test: out/soong/build.ninja is identical
Change-Id: Ie0135bdbd2df3258ef3ddb53e5f8fc00aa9b97f7
ParseGeneratorParameter delimits based on commas, not colons:
external/protobuf/src/google/protobuf/compiler/code_generator.cc
Bug: 72570104
Test: flash device with new build
Change-Id: I00042782a718d288ca6b0cb78f856d0fb223a926
This extends the minimal runtime dependency mutator to allow signed
and unsigned integer overflow diagnostics in static libraries and
binaries. This also enables the integer_overflow flag for static
libraries and binaries.
Note compilation will fail if the static library is a dependency
of a Make module that does not also have diagnostics enabled.
Bug: 66952339
Bug: 73283972
Test: make SANITIZE_TARGET{,_DIAG}=integer_overflow
Test: Enabled diagnostics in a static lib, saw results in logcat.
Test: Checked showcommands output for ubsan runtime library inclusion.
Change-Id: Ic52881a0f74cdcac0e4a15335df493b59b002ae5
Test: create reference dump for libjpeg; add exported function to libjpeg;
m -j libjpeg.vendor, build fails with helpful message.
Test: create reference dump for libjpeg; add exported function to libjpeg;
m -j libjpeg.vendor dist DIST_DIR=dist, build fails with helpful message.
Change-Id: Iae25374fe937a0cbe8a8ddf9e23c3bc1f62bbb2a
This commit changes the VNDK-SP dependencies check. With the commit,
VNDK-SP-Ext can link to non-VNDK vendor shared libs. This commit also
refines the "cc_test" so that more error handling cases are properly
tested.
Before this commit, VNDK-SP-Ext could not depend on vendor libs. It
was disallowed because there were no correct way to load vendor libs.
The fallback link had to specify the shared lib names. On the other
hand, adding "/vendor/${LIB}" to search paths will lead to
double-loading issue.
In aosp/595067, "allow_all_shared_libs" was added to bionic dynamic
linker. Now, we can link the "vndk" namespace to "sphal" namespace.
Thus, like VNDK-Ext, VNDK-SP-Ext can link to vendor libs now.
Bug: 77249955
Test: lunch aosp_walleye-userdebug && make -j8 # runs unit tests
Test: lunch aosp_sailfish-userdebug && make -j8 # runs unit tests
Test: Create a VNDK-SP-Ext, link to vendor libs, and run it.
Change-Id: I5511204539a22c998528111076f46756807faf29
Normally, when building with VNDK, platform modules are not allowed to
link against vendor libraries, because the ABI of the vendor libraries
are not guaranteed to be stable and may differ across multiple vendor
images.
However, the vendor public libraries are the exceptions. Vendor public
libraries are vendor libraries that are exposed to 3rd party apps and
listed in /vendor/etc/public.libraries.txt. Since they are intended to
be exposed to public, their ABI stability is guaranteed (by definition,
though it is up to the vendor to actually guarantee it).
This change provides a way to make a vendor lib as public by defining a
module of type 'vendor_public_library' with a map file that enumerates
public symbols that are publicized:
cc_library {
name: "libvendor",
proprietary: true,
...
}
vendor_public_library {
name: "libvendor",
symbol_file: "libvendor.map.txt",
}
This defines a stub library module named libvendor.vendorpublic from the
map file. `shared_libs: ["libvendor"]` is redirected to the stub library
when it is from the outside of the vendor partition.
Bug: 74275385
Test: m -j
Test: cc_test.go passes
Change-Id: I5bed94d7c4282b777632ab2f0fb63c203ee313ba
* The extra STL libs are:
libc++abi.a [needed for ndk_libc++_static]
libandroid_support.a [always needed in NDK r16]
libunwind.a [needed for ARM32]
* The existing STL-dependency logic in linkShared only applies to shared
libraries. By moving it to STL deps, the extra STL libs are linked into
both shared libraries and executables.
* Remove the ndk_prebuilt_library/ndkPrebuiltLibraryFactory module type,
which is unused now.
* Reuse the ndk_prebuilt_static_stl module type to describe the extra
static libraries that are linked with both the static and shared libc++
STLs.
Bug: b/73133405
Test: manual
Change-Id: I3f73e4f882d39e6efa470073bb4fc8c42dff8253
This CL changes PRODUCT_CFI_INCLUDE_PATHS to be included in all
product configs by default. To maintain the status quo, the sanitizer
logic has been modified to only respect this product config for Arm64
devices (where this was previously enabled).
Bug: 63926619
Test: m -j60 # the device still has CFI enabled thanks to the default
opt-in
Change-Id: Ia613aec545ad3e544dea41a6dbdb4112aef4afab
cfiExportsMap was reinitialized for every module, which caused
data races. Create the path from the string on each use
instead.
Bug: 77234104
Test: m nothing with race detector turned on
Change-Id: Ibca3149dcbe8a9d4d9f7ec6dd0b164697e7ae5cd
ExpandSourcesSubDir was calling SourceFileProducer.Srcs(), and then
doing:
moduleSrcs = append(moduleSrcs[:j], moduleSrcs[j+1:]...)
This modifies the backing store of the slice, which may affect the
original data stored in the SourceFileProducer. Make all Srcs
implementations return slice that points to a copy of the backing
array.
Test: m out/target/common/obj/PACKAGING/checkpublicapi-current-timestamp
Bug: b/76179848 b/76397326
Change-Id: I2432ce196984814daafc5aa9a2746e81de74494c
We omit vendor unavailable modules. This should not apply to llndk
libraries since currently, we do abi checks on their system variants.
Bug: 77101345
Test: m -j libc creates libc.so.lsdump
Test: m -j libjpeg.vendor creates libjpeg.so.lsdump
Test: m -j external/cblas still does not create libblas.so.lsdump
Change-Id: I5522c1cd471bfba8a1f632270ab2f167b4b17117
The space prevents "repo upload" from running on my machine:
$ repo upload .
[COMMIT 50e35bea87f0] Use Soong modules for the NDK's extra STL libraries
[FAILED] gofmt
FILES: ('cc/cc.go',) <standard input>
[FAILED] repohooks for platform/build/soong failed
FATAL: Preupload failed due to above error(s).
For more info, please see:
https://android.googlesource.com/platform/tools/repohooks/
Bug: none
Test: repo upload
Change-Id: I6b5de8f5d4edb38ca26a999e561df10262844b3d
* Suppress more noisy new warnings at global level.
* Add -no-pie to partial link .o files, with -r.
* Revert workaround of b/72706604, no need of
-Wl,-plugin-opt,-emulated-tls
* Filter out clang 7.0 unknown flag "-Wno-extended-offsetof"
Bug: 72706604
Bug: 72412006
Test: make checkbuild
Change-Id: I7ff45465c4bd771991f42b40f68dc35586045656
When building vendor modules with BOARD_VNDK_VERSION=current, the
API of the vendor modules will be current PLATFORM_VNDK_VERSION.
__ANDROID_API_FUTURE__ will be used as before if the version is a
CODENAME.
If BOARD_VNDK_VERSION is not "current", that means the VNDK version
of the vendor modules is BOARD_VNDK_VERSION.
Bug: 74833244
Test: Build and check boot.
Change-Id: I383c76a36101e39c70575b463880b52d3e9d90bb
The AIDL code generator can now include trace code in every generated
AIDL call. We don't want to enable this by default for all interfaces
for two reasons:
1) Potential performance overhead
2) For Java targets, the code depends on @hide APIs in android.os.Trace,
and I haven't found a clean way to determine whether a target is
allowed to use @hide stuff in the SDK. LOCAL_PRIVATE_PLATFORM_APIS
is almost it, but it's not set for the core framework (which is
exactly one of the things we'd like to trace).
Bug: 74416314
Test: verify correct code is generated when flag is set
Change-Id: Ic6ab8aee3e708207813da3e803911a844ff3e52b
Merged-In: Ic6ab8aee3e708207813da3e803911a844ff3e52b
This commit changes "current" to ctx.DeviceConfig.PlatformVndkVersion().
This change make it possible to place ABI dumps in
prebuilts/abi-dumps/vndk/${version} instead of
prebuilts/abi-dumps/vndk/current.
If BOARD_VNDK_VERSION is not specified or is equal to "current",
PLATFORM_VNDK_VERSION (either PLATFORM_SDK_VERSION or
PLATFORM_VERSION_CODENAME) will be chosen. Currently,
BOARD_VNDK_VERSION is always equal to "current" because both system and
vendor images are built in a single source tree.
In the future, if the system image and vendor images are separately
built in different source trees, BOARD_VNDK_VERSION will be set to other
values and we should respect that value.
Test: Create an ABI dump under
prebuilts/abi-dumps/vndk/${PLATFORM_VNDK_VERSION}, make some ABI
breakages, and see whether the checks are working.
Bug: 76036094
Change-Id: I6242e0c71ebd9acd5c4a3497c67539cb3e6663f3
If a generated renderscript header is referenced by a cpp file
clang will insert a depsfile dependency on the header. If the
generator rule does not include the header as an output, ninja
will not consider the cpp file dirty after the generator runs,
and not rebuild it. On the next build, it will see that the
timestamp of the generated header is new and consider the cpp
file dirty.
Mark the header files as implicit outputs of the generator.
Bug: 75982985
Test: m libstagefright_mediafilter
touch frameworks/av/media/libstagefright/filters/saturationARGB.rs
m libstagefright_mediafilter
m libstagefright_mediafilter
Change-Id: I3ac899322fe9cae2a6e4171092740339ef844494
* This is a workaround of llvm LTO bug.
* Keep using current NDK include directory path.
NDKTriple is default ClangTriple but kept as GccTriple for arm device.
Bug: 72619014
Test: make checkbuild
Change-Id: I5dc63c99760325c60bc2da98fd6a3125cef7267d
Allow developers to enable ThinLTO caching on their local build by
setting USE_THINLTO_CACHE environment variable.
This significantly speeds up incremental ThinLTO builds by caching
intermediate results. An incremental ART build (by `touch
compiler/compiler.cc`) reduced from 01:32 to 00:45.
We put the cache under out/soong/thinlto-cache as the cache can be
valid across different targets, and allows us to set a global cache
pruning policy.
The current cache policy is sufficient for current list of
ThinLTO-enabled projects, we can tune it up/down in the future.
Test: USE_THINLTO_CACHE=true m
Bug: 62839002
Change-Id: I4838fe833a1d8fbc73d743da6d19e3698dc89600
Instead of open-coding the logic of whether there is one, or where to
find it.
Test: diff out/soong/build.ninja without dist
Test: diff out/soong/build.ninja with dist specified
Change-Id: Ia3f1ef335e2d6e2175343338d04867d778a50300
So that the Path and similar functions can be used directly, without
manually adding something like configErrorWrapper (it just uses it all
the time now).
Test: out/soong/build.ninja is identical
Change-Id: I8cb524b09a84d0b8357d7c3501c71270af411e17
Instead of whichever xz happens to be on the PATH.
Bug: 36130900
Test: prebuilts/build-tools/build-prebuilts.sh
Change-Id: Ic9b66062ca7947f0eca9d839c0864da492d4ef71
Bug: http://b/74395273
Update profile-search logic to look for profile files named
<profile_file>.<arbitrary-version> in the profileProject and use that if
available. This works around an issue where ccache serves stale cache
entries when the profile file has changed.
Test: touch toolchain/pgo-profiles/art/dex2oat.profdata.foo and check it
gets used.
Change-Id: I3de095c263b952cede25af7d0b63dc7280914d76
Test: Add proto.canonical_path_from_root: true in a cc_defaults
Test: Add proto.canonical_path_from_root: true in a java_defaults
Change-Id: I9ddfc8af0025705a34b6e487225f1f98915054c3
Adds Soong support for -fsanitze-minimal-runtime when using
the integer overflow sanitizers. This makes the crashes due to these
sanitizers less mysterious.
Bug: 64091660
Test: Compiled and checked the generated compiler commands
Test: Checked program that overflows for the abort reason
Change-Id: Ieeceaf6c35c8371592952d3b8b977aefc11601c5
Vendor binaries cannot directly link with vndk libraries which are not vendor
available. We don't need them to be abi stable.
Bug: 66301104
Test: cd external/cblas; touch src/cblas_cher.c; mm -j64 does not
produce an lsdump for libblas, since it isn't vendor available,
even though it is vndk_enabled: true.
Change-Id: Ib2e1eaa06c4dc2e05623bd7b9aa3a83010f76bd1
Allow native modules to specify use_version_lib, which will make
an android::build::GetBuildNumber() function available. For host
builds, the function will return the build number at the time that
the module was linked. For device modules it will return the
value of the ro.build.version.incremental property.
Bug: 71719137
Test: build_version_test
Test: m build_version_test && touch build/make/core/Makefile build/soong/cc/libbuildversion/tests/build_version_test.cpp && m build_version_test shows different build numbers for binary and library tests.
Change-Id: I6f7d40b7574bb8206866c4e39bad9c710c796e32
asan variants can sometimes have extra exported symbols (this is a vndk
extension violation). Since asan variants are only used for testing,
don't do abi diffs for them.
Bug: 66301104
Test: Without the change, libc++.so.lsdump exists for the asan variant;
with the change, it does not.
Test: make -j64.
Change-Id: I61a4c7a3e9aa0028a54ad0ca8715e8c77aebad94
Bug: http://b/73127367
Bug: http://b/68659946
These new prebuilts have fixes for the bugs mentioned above.
Test: Build and verify that issues in both bugs above are fixed.
Change-Id: I410a43b70920d509118f8e6fdb012ef6b078d657
This is needed so that we can embed the paths of lsdump files in a file,
in order to make lookup faster, while creating reference dumps.
Test: mm -j64 in external/libjpeg-turbo for aosp_arm64_ab;
out/soong/make_vars-aosp_arm64_ab.mk contains
SOONG_LSDUMP_PATHS := <list of lsdump paths>
Change-Id: I04608429e1add307cc1ee79d2f0c348fb041613c
GlobFiles had allowed results to be anywhere in the source tree,
restrict it to results within the current module directory.
Then use it for ExpandSources and other places where we only want files.
This fixes using '*' in cc_test's `data` property, which can only
support files.
The only thing this changes today is that java_resource_dirs and
java_resources no longer pass directories to soong_zip's -f argument.
core-libart previously added some icu directories, now it only passes
files.
Bug: 71906438
Test: only expected changes in out/soong/build.ninja
Test: add data: ["**/*"] to a cc_test, build successfully
Change-Id: Iff1bd8c005a48e431c740706d7e23f4f957d8b1d
Historically, we've always passed '-I .' as the first argument to
protoc, essentially treating all proto file package names as their full
path in the android source tree. This would make sense in a monorepo
world, but it makes less sense when we're pulling in external projects
with established package names.
So keep the same default (for now), but allow individual builds to opt
into using local paths as the default names with
'canonical_path_from_root: false'. A cleanup effort and/or large scale
change in the future could change the default to false.
As part of this, run protoc once per input proto file, since the flags
may need to change per-file. We'll also need this in order to specify
--dependency_out in the future.
Bug: 70704330
Test: aosp/master build-aosp_arm.ninja is identical
Test: aosp/master soong/build.ninja has expected changes
Test: m
Test: Build protobuf test
Change-Id: I9d6de9fd630326bbcced1c62a4a7e9546429b0ce
While the rule may really need all of the generated header files to
exist, only one of them (per genrule task) needs to be in the dependency
list, since the rest are essentially aliases.
This brings an AOSP aosp_arm-userdebug out/soong/build.ninja file from
372MB to 156MB, with equivalent functionality. The Android-aosp_arm.mk
file is reduced from 11MB to 6.5MB.
Bug: 73745773
Test: diff out/soong/build.ninja
Test: diff out/soong/Android-aosp_arm.mk
Test: rm -rf out; m
Change-Id: If17377666292cc20957417fc4c3cd52f98971d0c
Disabled VNDK libraries (due to arch mismatch) are no longer emitted to
make variables VNDK_CORE_LIBRARIES and LLNDK_LIBRARIES
Bug: 7456955
Test: choosecombo to aosp_walleye and m -j
Find libclang_rt.ubsan_standalone-<arch>-android in VNDK_CORE_LIBRARIES.
Only aarch64 and arm are found.
Change-Id: Iaa134d07513e39390fe34a31fdfe2e327b190996
This reverts commit 606e9de344.
Reason for revert: <try to fix the broken build yesterday>
Change-Id: I2963b9af63c7c7398159e5e9a1e448266e1c81d5
Test: unittest
Such optimisations may significantly increase the binary size when
compiler heuristics are off. Disabling these helps cut down the
binary sizes with negligible decrease in performance, but allows us to
be more comfortable enabling LTO across various projects.
Test: m
Test: dex2oat, hwui, skia benchmark
Bug: 62839002
Change-Id: Id63e8dd295df2972f76ae4e29ee367080fff8429
LLVM-AR does not allow passing --plugin options more than once. The
--plugin ARFLAGS that lto want to add, may already exist if sanitizer is
also turned on.
Fixed this by adding a new bool Flags.ArGoldPlugin. Set this variable to
true whenever LLVM gold plugin is needed for ArFlags. In function
TransformObjToStaticLib(), add this option to arFlags using global value
${config.LLVMGoldPlugin} if the bool value is true.
Bug: http://b/73160350
Test: build the image with make and succeeded.
Change-Id: I62785829b0a4b663225926e4aed98defc1b6da2c
(cherry picked from commit 4917049f6e)
This variable can be set in BoardConfig.mk to specify a list of
additional paths that contain PGO profiles. These directories are
searched after the predefined paths in soong/cc/pgo.go while finding
PGO profiles.
Test: Set this variable in a BoardConfig and verify that such profiles
are found and that these paths are searched after the predefined paths
in soong/cc/pgo.go.
Change-Id: I0bb9523de614d0f23aba8d51c887d8fc8f41c993
-fsanitize=cfi requires LTO. Normally, the Clang driver automatically
enables emutls, but it's broken with -flto, so work around it by passing
-Wl,-plugin-opt=-emulated-tls.
See https://buganizer.corp.google.com/issues/72706604#comment15
Enable the workaround with ctx.Device(), which is the same condition used
in the existing I18acac41aac885fc6635fbd55f96ba7c845eb5e7 workaround.
Bug: b/72706604
Bug: https://github.com/android-ndk/ndk/issues/498
Test: m libaudioflinger (on internal master); \
readelf -sW symbols/system/lib64/libaudioflinger.so | grep tlNBLogWriter
check that __emutls_t.tlNBLogWriter exists
check that no symbols of type TLS exist
Change-Id: I2cf65574c52476843cc017ee176a7d6777e2ce0b
We can simply revert this patch to work on these warnings independently.
Bug: http://b/72331526
Bug: http://b/72331524
Bug: http://b/72330874
Test: Build
Change-Id: I9cec51578a1c6de33fb6f832c758b9916c61b735
