This CL is scoped to cc_* module types. With trunk stable, we will have
multiple prebuilts of the cc modules in
prebuilts/module_sdk/art/<v>/host-exports/, and this CL introduces a
mechanism to use apex_contributions to select a specific versioned
prebuilt when building.
If a soong module is selected using apex_contributions, all rdeps will
get that soong module, which includes
- rdep soong modules which might be depending on it via Android.bp
- Soong's rule builder HostToolPath API
Implementation details: Create a new source_module_name property to
identify the root module. rdeps referring to the root module will get
redirected if necessary. This property also
becomes the stem, if `stem` is not set explicitly.
Bug: 322175508
Test: Added a unit test
Change-Id: Ic8725602c81999621fcb33ce2a57fe4b9751baa8
Add the dependency on the fdo_profile module in afdoTransitionMutator
and read the provider in GenerateAndroidBuildActions.
Bug: 319288033
Test: afdo_test.go
Change-Id: Ied8fd7b52d5694a3691652318e87b8fe14dda126
Convert afdoDepsMutator and afdoMutator to a TransitionMutator as a
step towards variants-on-demand.
Bug: 319288033
Test: afdo_test.go
Change-Id: Ib05845455ccf43a07b3915a0d7b0a95896062f13
Convert orderfileDepsMutator and orderfileMutator to a TransitionMutator
as a step towards variants-on-demand.
Bug: 319288033
Test: orderfile_test.go
Change-Id: I27df65b7264a5b059a900e7ea04c18dfb2787a84
Convert ltoDepsMutator and ltoMutator to a TransitionMutator as a
step towards variants-on-demand.
Bug: 319288033
Test: lto_test.go
Change-Id: I2c9af73fd526adf58ff626831ababea466338205
Convert coverageMutator to a TransitionMutator as a step towards
variants-on-demand.
Bug: 319288033
Test: coverage_test.go
Test: treehugger coverage builds
Change-Id: Ic50c0040dea8b42c36b5d784221daa00b7b0d379
For chromebooks, we need a new build variant whose target arch is
goldmont, but without SHA and XSAVES support (b/314243939#comment21)
Bug: 314243939
Test: Build an image with TARGET_ARCH_VARIANT:=goldmont-without-sha-xsaves
Change-Id: If73660b515b443d5c138ca367fa3d1c6f18485b5
Although vendor modules can be built bundled, they are technially not
part of the "platform", as it cannot access platform libraries (system
partition) directly.
They can only link to a restricted set of platform libraries (LLNDK),
and access restricted set of APIs.
We used to derive the LLNDK API level from the VNDK version. However
after VNDK deprecation, there is no "VNDK version" anymore.
Instead we would just derive the value from platform SDK version:
* If building an in-development build, build vendor modules against the
in-development "current" API level.
* If building a REL / Final build, vendor should target the latest
stable API.
Bug: 320423828
Test: go test
Test: presubmit
Test: build and boot
Change-Id: I2c5ef6530e9046b2dcc282bc1f020d8a505eab15
No-one ever fixes unused-* or deprecated-* warnings in external/ (nor should they!).
-Wno-deprecated-enum-enum-conversion should probably have been added
next to -Wno-deprecated-anon-enum-enum-conversion when it was added,
since it's a major spam contributor, and equally unlikely that anyone
will clean them up.
-Wno-deprecated-dynamic-exception-spec is a weird special case because
it should be irrelevant after our next libc++ update, but in the
meantime -- because libc++ is used everywhere -- this is a massive spam
contributor that no-one on Android cares about because we don't use
exceptions (and the people seeing this can't fix libc++ anyway).
Test: local clean builds
Change-Id: I098202337e9e5026c4c5215dbf5a2abf1fbbdca8
The previous patch was insufficient. Craig Topper explains:
-fno-vectorize only disables the loop vectorizer
-fno-slp-vectorize only disables the SLP vectorizer
The backend can also use vector instructions for memcpy/memset or
combining multiple scalar loads and/or stores. That is independent
of -fno-vectorize.
-mllvm -vector-bits-min=0 will disable any use of fixed vectors. And
will make attribute(vector_size) get scalarized.
-mno-implicit-float will disable both vectorizers and
prevent the backend for using vectors for memcpy/memset
oor multiple scalar loads/stores. It will not affect
attribute(vector_size). -mno-implicit-float also prevents scalar
floating point instructions from being used for anything that didn’t
use float/double/_Float16/etc. type in source, but I don’t think that
happens on RISC-V today. 32-bit X86 can use a 64-bit x87 FP load as an
atomic load for uint64_t, for example. Basically -mno-implicit-float
is supposed to prevent the compiler from using FP or vectors when the
source doesn’t explicitly use FP or vectors.
So -mno-implicit-float was what we were actually looking for here. I've
done a clean build with this change, and see only the expected
(hand-written assembler) vector code in bionic, and the ART
ClassLinker::LoadClass() issue is gone too. As far as I can tell, the
remaining vector code is all deliberate in that sense.
We may still end up back here again, to change "gcv" to "gc", but that
still requires some code changes just to build, and still makes it less
obvious that this is just a temporary workaround for a qemu bug
(specifically https://gitlab.com/qemu-project/qemu/-/issues/1976).
Bug: http://b/320416684
Test: objdump
Change-Id: Ibd104e4289d6d1aaf441efa0440fedc90e3da29a
* changes:
Add script to run Soong tests with go tools
Disable TestVariantSingletonModule when go test -short is used
Fix data race in propagateRROEnforcementMutator
Fix data race in finder_test.go
Fix data race in dex_bootjars
Fix race CommonGlobalCflags when running tests in parallel.
Fix data race in snapshot singletons when running parallel tests
Until https://gitlab.com/qemu-project/qemu/-/issues/1976 is fixed, this
is just causing too much confusion and wasted time.
Removing V from the architecture string has the same effect, but
requires some code changes (to be able to compile code that's
_explicitly_ using V), and makes it less obvious that this is just a
temporary workaround for a qemu bug.
Bug: http://b/320416684
Test: objdump
Change-Id: Iec7d63ab227b31f82f530dabbdc4f3aa54277f04
Clone the commonGlobalCflags global variable when modifying it so that
it gets a unique object per context.
Test: go test -race ./...
Change-Id: Ia4b91c40939d4b4d338bd11e4d4f523521874548
This change is to fix more misc issues to enable Soong without VNDK
version. This change contains
* Update properties in generated Android.mk
* Update VNDK APEX build to work without VNDK version
Bug: 316829758
Test: AOSP Cuttlefish build succeeded
Change-Id: I10f3c798299afe2d539ec3426b8e2b6068a158f6
Generated codes like cpp code from *.ll for *.yy files always run
clang-tidy, when generated code has warning, where is no way to fix it.
So, disable clang-tidy for generated code.
Bug: 162909698
Test: go test android/soong/cc
Change-Id: I0fee137d6170ef4bf6cf641abad572e448aceaa1
Image variants for CC modules were covered to work without VNDK version
from previous commit, but CC genrule has its own image variant rule, so
it should be also updated to generate vendor / product image variant
without VNDK version.
Bug: 316829758
Test: m nothing --no-skip-soong-tests passed
Test: AOSP cuttlefish build succeeded
Change-Id: I425dd425efcc57c7ff8f9964b303ad6f539c3b57
Since VNDK in the main branch will not be finalized, this commit
deletes the build rules that generate the ABI dumps.
Test: make
Bug: 314010764
Change-Id: I2bbab7612d8cbc27312dc4a3c47a0592177918d8
UseVndk is a function to check if the module can use VNDK libraries, but
this function was also used to check if the module is installed in the
treblelized partition (vendor or product). As of VNDK deprecation,
UseVndk funtion will return false even when the module is installed in
vendor / product partition, so we need a separated function to check
this. This change introduces a new function 'InVendorOrProduct' which
replaces UseVndk based on its usage.
Bug: 316829758
Test: m nothing --no-skip-soong-tests passed
Change-Id: Ic61fcd16c4554c355f6005894a4519b044b27fe5
Current CC/Rust Image variations are generated with target VNDK version.
However, this is no longer valid if VNDK is deprecated. This change
generates image variation without version ("vendor", "product") if VNDK
is deprecated.
Bug: 316829758
Test: m nothing --no-skip-soong-tests passed
Test: aosp_cf_x86_64_phone build succeeded
Change-Id: I2387ed8a2632bfd9462621f882a947695ae1653d
libz is a stub library, but needs to be available to runtime apex
because it gets statically linked into bionic linker
Bug: 281077552
Bug: 277651159
Test: m nothing
Change-Id: I04f6f13768d8f9c160ce84202e2003b195176355
LOCAL_STATIC_LIBRARIES, LOCAL_WHOLE_STATIC_LIBRARIES,
and LOCAL_HEADER_LIBRARIES are only exported to Make
so that it can generate module-info.json. Export
ModuleInfoJSONProvider from cc modules so that Soong can
generate the module-info.json entries, and remove the
properties from the generated Android.mk. This will prevent
Kati reanalysis when making some Android.bp changes.
Bug: 309006256
Test: Compare module-info.json
Test: Compare Kati's build.ninja
Change-Id: I6660f6802b9cea46eed553cac12f09a373eeb019
This wasn't needed when fuzzers had their own
lib directories, but because the lib directories
were combined, this caused interactions because
system and vendor libraries were mixed in the
same folder, sometimes overriding each other.
Comments are left with some details about
future issues. Bug has a series of recommendations
to prevent issue again.
The main goal of this patch is to get the fuzzers
working, not fix every issue here.
Bug: 307611931
Test: run vendor fuzzer and check lsof:
android.hardwar 22719 root txt REG 254,94 4350056 3455 /data/fuzz/x86_64/android.hardware.drm-service.<name>.aidl_fuzzer/vendor/android.hardware.drm-service.<name>.aidl_fuzzer
android.hardwar 22719 root mem REG 254,94 4350056 3455 /data/fuzz/x86_64/android.hardware.drm-service.<name>.aidl_fuzzer/vendor/android.hardware.drm-service.<name>.aidl_fuzzer
android.hardwar 22719 root mem REG 254,94 1047232 3464 /data/fuzz/x86_64/lib/vendor/libbase.so
android.hardwar 22719 root mem REG 254,94 4027304 3460 /data/fuzz/x86_64/lib/vendor/libcrypto.so
android.hardwar 22719 root mem REG 254,94 4021632 3467 /data/fuzz/x86_64/lib/vendor/libbinder.so
android.hardwar 22719 root mem REG 254,94 329408 3466 /data/fuzz/x86_64/lib/vendor/libcutils.so
android.hardwar 22719 root mem REG 254,94 508968 3463 /data/fuzz/x86_64/lib/vendor/libutils.so
android.hardwar 22719 root mem REG 254,94 592944 3465 /data/fuzz/x86_64/lib/vendor/libclang_rt.ubsan_standalone-x86_64-android.so
android.hardwar 22719 root mem REG 254,94 2328616 3462 /data/fuzz/x86_64/lib/vendor/libc++.so
and run a system fuzzer
servicemanager_ 27878 root txt REG 254,94 609736 67408 /data/fuzz/x86_64/servicemanager_fuzzer/servicemanager_fuzzer
servicemanager_ 27878 root mem REG 254,94 609736 67408 /data/fuzz/x86_64/servicemanager_fuzzer/servicemanager_fuzzer
servicemanager_ 27878 root mem REG 254,94 4390560 4074 /data/fuzz/x86_64/lib/libvintf.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 11584 4086 /data/fuzz/x86_64/lib/libdl.so
and run a system fuzzer
servicemanager_ 27878 root txt REG 254,94 609736 67408 /data/fuzz/x86_64/servicemanager_fuzzer/servicemanager_fuzzer
servicemanager_ 27878 root mem REG 254,94 609736 67408 /data/fuzz/x86_64/servicemanager_fuzzer/servicemanager_fuzzer
servicemanager_ 27878 root mem REG 254,94 4390560 4074 /data/fuzz/x86_64/lib/libvintf.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 13872 4077 /data/fuzz/x86_64/lib/libvndksupport.so
servicemanager_ 27878 root mem REG 254,94 11584 4086 /data/fuzz/x86_64/lib/libdl.so
Change-Id: I91a51d2f1cb537cfaae8379998078188f2b10a98
Overriding android.ModuleContext's implementations of *Specific()
methods in cc.moduleContext and then passing that back to
android.PathForModuleInstall to affect the install path causes
problems if android.ModuleBase.GenerateBuildActions also tries
to call android.PathForModuleInstall directly with the
android.ModuleContext as it gets a different result.
Add InstallIn* methods to the android.Module interface, implement
default versions in android.ModuleBase, and override them in
cc.Module and rust.Module. Use them in android.PathsForModuleInstall
to allow the module to customize the behavior directly.
Test: TestInstallPartition
Change-Id: I7840e07eae34ac4f4d3490b021143d5f33a83626
Its target arch is goldmont, but without xsaves support.
This ensures efficient execution on a broad range of Intel/AMD CPUs used
in Chromebooks, including those lacking xsaves support.
(e.g. Kaby Lake, Gemini Lake, Alder Lake and AMD Zen series)
Bug: 314243939
Test: Build an image with TARGET_ARCH_VARIANT:=goldmont-without-xsaves
Change-Id: I9963cc3356394815f068fc998a9708bdb30c8266
Convert all of the callers of ModuleProvider/ModuleHasProvider to use the
type-safe android.SingletonModuleProvider API.
Bug: 316410648
Test: builds
Change-Id: I6f11638546b64749e451cebbf33140248dc1d193
Convert all of the callers of OtherModuleProvider/OtherModuleHasProvider
to use the type-safe android.OtherModuleProvider API.
Bug: 316410648
Test: builds
Change-Id: Id77f514d68761a262d9ea830a601dbed804bbbe5
Convert all of the callers of Provider/HasProvider to use the type-safe
android.ModuleProvider API.
Bug: 316410648
Test: builds
Change-Id: I73479de1625fa2865b6c73444cd477e50d56dc5a
Convert all of the callers of SetProvider to use the type-safe
android.SetProvider API.
Bug: 316410648
Test: builds
Change-Id: If58f4b5355264ddab2045bc3591a4eac19cd58fc
Convert all of the callers to NewProvider and NewMutatorProvider
to use a generic type parameter instead of an example object.
Bug: 316410648
Test: builds
Change-Id: Ic9cdafc87336e26730d3fd596df05de0e7267542
Using generics for the providers API allows a type to be associated
with a ProviderKey, resulting in a type-safe API without that doesn't
require runtime type assertions by every caller.
Unfortunately, Go does not allow generic types in methods, only in
functions [1]. This prevents a type-safe API on ModuleContext, and
requires moving the API to be functions that take a ModuleContext as
a parameter.
This CL creates the new API, but doesn't convert all of the callers.
[1] https://go.googlesource.com/proposal/+/refs/heads/master/design/43651-type-parameters.md#no-parameterized-methods)
Bug: 316410648
Test: builds
Change-Id: I3e30d68b966b730efd968166a38a25cc144bd6de
clang silently ignores such things rather than reporting them as errors.
With this change, verbose.log.gz says android10000 instead, as expected.
Bug: https://buganizer.corp.google.com/issues/315788463#comment24
Test: treehugger
Change-Id: I0d92b0747912f28065de0d2dc6b4d33d1c2c9857
Remove the ConvertWithBp2build implementations from all the module
types, along with the related code.
Bug: 315353489
Test: m blueprint_tests
Change-Id: I212672286686a318893bc7348ddd5a5ec51e77a7
Passing the list of all transitive aconfig files to Make causes extra
Kati analysis runs when dependencies are changed in Android.bp files.
Since Make is going to merge them anyways, merge them per-module and
pass a single aconfig file to Make for each module.
Fixes: 313698230
Test: m out/target/product/vsoc_x86_64/system/etc/aconfig_flags.pb
Change-Id: Ifde4826bc93bc06e40338f72b4cb39eed26ca08d
Currently NDK ABI monitoring is enabled by switching boolean flag in
soong code. When this flag is switched, it will enable ABI monitoring in
all release configurations at once.
This change moves the flag from soong code to the trunk-stable flags
framework ("build/release/build_flags.scl") that allows to enable it
only in specific configurations like "trunk-staging".
Test: TH
Bug: 156513478
Change-Id: Ibb810a834c0d756b22782b1a2d8ec5a5e575a4b3
This is part of a larger change to update the flag
PRODUCT_PAGE_SIZE_AGNOSTIC to PRODUCT_NO_BIONIC_PAGE_SIZE_MACRO.
This is to help clarify that this flag doesn't mean the device
will actually work end-to-end with any page size.
Bug: 312541564
Test:
```
source build/envsetup.sh
lunch aosp_cf_arm64_phone_pgagnostic-trunk-userdebug
m
cat out/soong/build.aosp_cf_arm64_phone_pgagnostic.ninja | grep __BIONIC_NO_PAGE_SIZE_MACRO
```
Change-Id: I33033876d0d4a276d1bb962d40315b71a3968c66
When generating vendor/etc/linker.config.pb, STUB_LIBRARIES was used.
Because it lists *all* libraries with stubs regardless of its
install location (vendor or system), using it for vendor
linker.config.pb doesn't make sense. Instead, use vendor-specific
STUB_LIBRARIES, which is named SOONG_STUB_VENDOR_LIBRARIES.
Bug: 313806237
Test: m (aosp_cf_x86_64_phone)
Test: $OUT/vendor/etc/linker.config.pb not listing libz.so
Change-Id: Icd0aaf92d9630c07f58c4739a9f0ac713516db43
Afdo creates more variants to build, and optimizing LTO is costly to
perform. Turn off these two optimizations for eng builds for a faster
build speed.
This reduces total C/C++ invocation time for aosp_arm64-eng by 6.28%.
Test: presubmit
Bug: 307753064
Change-Id: Ibac4269c66a64e896dba2074b607d71a2da37546
This is to handle libz special case. libz has stubs but not an LLNDK.
So, libz.vendor should be treated as non-stub-providing libraries and
Vendor APEX should bundle it if it's used by its contents.
libz will set no_stubs for vendor/product variants.
Bug: 313806237
Test: go tests ./apex/...
Change-Id: I10759d7073838909126f8bfe87654f11aa02fd32
The next CL will need a TestContext parameter in
ContentFromFileRuleForTests in order to retrieve the file rule contents
from the Config. Add it and update all the tests that use it in order
to simply review of the next CL.
Bug: 306029038
Test: go test ./...
Change-Id: Ia4b4c9854017ea3472fa2f8ba42cf7f72720496e
By default clang assumes that on riscv64 jumps are really expensive.
That's probably not true for the kind of SoCs we're dealing with, but
more importantly (see the bug) it causes clang to do loads from the
right hand side of a `&&` before the left hand side has been evaluated.
This found one latent bug in libcore, and it doesn't seem like the best
use of anyone's time to have to chase similar issues if they're going
to be similarly latent for riscv64 when we get clang's default fixed.
Bug: https://github.com/google/android-riscv64/issues/124
Test: treehugger
Change-Id: I640f1b43ea3d2452366ab86e97a9189fa9f5326c
To generate module-info.json in Soong for b/309006256 Soong needs to
know the test data paths. Moving test data installation into Soong will
also help later for test suite packaging.
Add ModuleContext.InstallTestData that installs the files listed in a
[]DataPath alongside the test. The files will also be passed to Make
to allow it to continue packaging them into the test suites for now.
Update the module types that are producing LOCAL_TEST_DATA entries
in their Android.mk files to go through InstallTestData instead.
Bug: 311428265
Test: atest --host toybox-gtests --test-timeout=120000
Change-Id: Ia8b964f86e584ea464667fd86a48d754d118bead
cc modules can use __ANDROID_VENDOR_API__ to read BOARD_API_LEVEL
that is the API level of the vendor surface.
Bug: 313822931
Test: check ninja commands to have -D__ANDROID_VENDOR_API__=<version>
Change-Id: Iceb5918cbfce0f24929d02d7e3caf1e9942b88e8
This can be used to add apex-specific shared_libs to a CC module. It
would be nice if we could have apex.shared_libs. But it would make code
much harder to follow because we need to record the shared_libs list as
"exclude list" for non-apex anyway.
Bug: 312510312
Test: m libhardware libhardware.vendor libhardware.vendor.com.google.cf.ir
# check DTNEEDED for libapexsupport
# only apex-variant should have the dependency
Change-Id: I56a3dc280127d8ba44337707444ea226a49ccf0f
For LTO compilation, we do not need MLGO flags in cflags, since codegen
happens during link only.
Fixes compiler warnings for unused command line argument.
Test: presubmit
Change-Id: I361e4292b10a3582fd5c69fa7b5678c654b44a0f
This Cl sets testModule field to True in cc.NewTest(). This will cover "cc_test", "cc_test_host" and "art_cc_test".
Change-Id: I4a8db86835b195db34fd9f86560e7bf9321fbd7d
Provider added for the following test modules in this change: art_cc_test, cc_benchmark, cc_fuzz, cc_test, cc_test_host, rust_test,and rust_test_host.
Bug: 296873595
Test: Manual test
Change-Id: I815680529bcbecacb3a2bdb8f3746053afdee48c
Fix various issues that prevent building or running tests with the
Go tools:
Remove testing/test_spec_proto/go.mod. There's already a go.mod at
the top level, and adding testing/test_spec_proto/go.mod breaks
compiling with the Go tools in IntelliJ or from the command line.
Fix go vet issues when trying to print []DataPath.
Fix aconfig_conversion_test.go and add it to Android.bp so it runs
in presubmit.
Test: go test ./...
Change-Id: I18af2f05631446fe507744a2951e4b02c7ab59be
We should update gnu++2a to gnu++2b too, but not in this CL.
Bug: 311052584
Test: treehugger
Test: m checkbuild
Change-Id: I3637444d74868b3a29b425fd749ecbbe1a45fbf7
This reverts commit febf8b9bdf.
We are still seeing flaky performance alerts from microbenchmarks even
with this change, and this has caused binary size / memory usage
increase. Let's add this flag back and observe whether performance
regression reports become more frequent.
Change-Id: I0fec1296b1f49b21b59d81add677f6a926ee88dc
To simulate 16k page size in x86-64 cuttefish, it is necessary
to compile the shared libraries and binaries with max-page-size greater
than 4096.
Bug: 309816695
Test: source build/envsetup.sh
lunch aosp_cf_x86_64_phone_pgagnostic-trunk-userdebug
m
Change-Id: I32670ef2c25fdcefec11bd07ba41cd0ea96c92bd
To simulate 16k page size in x86-64 cuttefish, it is necessary
to compile the shared libraries and binaries with max-page-size greater
than 4096.
Bug: 309816695
Test: source build/envsetup.sh
lunch aosp_cf_x86_64_phone_pgagnostic-trunk-userdebug
m
readelf -l out/target/product/vsoc_x86_64_only/system/bin/toybox
Change-Id: I38e4449a7826a2eca0c1a1cccc916af21f0b3a93
The MLGO model we embed in the Android Clang is trained on ARM64
ThinLTO artefacts. Applying the model across architectures does not
necessarily translate to performance improvement. Therefore only enable
the optimisation on ARM64 ThinLTO for now.
Test: presubmit
Bug: 293827654
Change-Id: Ie64a65c11191cf700a463637a0746c8470d3aa3c
In a trunk stable branch, the ABI-monitored libraries under different
release configurations share the source code. The exported symbols are
filtered by version scripts and API level. header-abi-linker needs to
take the API level as a parameter.
Test: make
Bug: 308192597
Change-Id: I4b55b7ec023451b3a4cd1da308a1de10842d51d2
Also add -Wno-everything as suggested by Pirama
Bug: b/308179696
Follow up: https://b.corp.google.com/issues/308179380
Change-Id: Iaa73f2b52363562b2391346d8b2e5dc73bab4f2d
Product modules install required vndk modules in /product/lib[64].
But still the product modules can link to the llndk from system
partition using the list in /system/etc/llndk.libraries.txt.
To install /system/etc/llndk.libraries.txt file, use
llndk_libraries_txt singleton without versioned name.
Until VNDK is fully removed, the vndk apex must install the versioned
llndk.libraries.<ver>.txt file to the vndk apex.
Use llndk_libraries_txt_for_apex singleton for this purpose.
Test: lunch cf_x86_64_phone-next-userdebug; m
Test: atest GtsVndkDependencyTestCases
Bug: 299867815
Bug: 302255959
Change-Id: I058616081fe31597a9b76b772b69927807436896
This warning is triggered even when initializers don't have any
side effects, so it very rarely surfaces actual bugs/mistakes.
99% of the time it complains about a purely cosmetic issue.
Disable it by default in commonGlobalCFlags, which still allows
individual modules to opt into it.
Bug: 145210666
Test: presubmit
Change-Id: I671b5c457ee3d92b9cc37f020ba54e01e1a04f59
More might be added later when we have an allowlist.
Bug: 305998969
Test: b build with the change in ag/25094394
Change-Id: I92df843ac81f4e33ce7506db61c989a380015975
* changes:
Revert "Decompress debug sections when creating debug info"
Revert "Compress debug info in clang builds with zstd"
Revert "Use zstd to compress debug info in cc libraries"
Revert "Use zstd to compress debug info in rust libraries"
* changes:
Use zstd to compress debug info in rust libraries
Use zstd to compress debug info in cc libraries
Compress debug info in clang builds with zstd
Decompress debug sections when creating debug info
We can save a significant amount of disk space by compressing
debug info in `lld` cc builds. This change configures a build
flag to instruct lld to compress debug info when building cc
libraries. Rather than adding a global config we add this to
each target individually to avoid linker failures on windows
cross build targets.
Test: m
Bug: 305277519
Change-Id: I4ab3d47fb0de7e31a39fb671cccde1acee3a2018
We can save a significant amount of disk space by compressing
debug info in `clang` builds. This change configures a build
flag to instruct clang to compress debug info.
Test: m
Bug: 305277519
Change-Id: Ieeaa655452d847e83ebc6bec0864e47c46f4c6fb
This reverts commit 4e28d0eda2.
Reason for revert: Reverting now that we should have sufficient data from Crystalball https://b.corp.google.com/issues/306023954
Change-Id: I9f92c203eda309956ec192432ba0063eb11f59e2
This reverts commit 07c362bf7f.
Reason for revert: DroidMonitor: Potential culprit for Bug 304839693 - verifying through ABTD before revert submission. This is part of the standard investigation process, and does not mean your CL will be reverted.”.
Change-Id: Iafb894510c48af2dd8b015d7264a031ff9348884
This enables sharing of platform constraints between platform and kernel builds.
Bug: 304594734
Test: treehugger
Change-Id: I48091eeaf2e529c3f180ba056da3935911868b14
FSAFDO (Flow Sensitive AutoFDO) is an improvement over AutoFDO. It
attempts to capture and represent the flow-sensitive profile in LLVM
optimization passes with the mechanism to expose and iteratively
reload/match the profile with better precision.
Enable this for all AFDO projects so that we can capture data from the
field, and then evaluate the performance gain.
Test: presubmit
Bug: 304282732
Change-Id: I0de6b4071e4f0a75500f3ffd5171a7be5117ec86
This reverts commit 80b54d2502.
Reason for revert: caused hello_world_test to not be converted, but it's a test that builds/passes
Bug: 300117121
Change-Id: I805cfb336b7f58a5a1e295cade16be4c471e2ed5
(This CL is a cleanup, and should be a no-op)
Currently we support three variations of cc api stubs.
1. publicapi stubs, i.e. ndk stubs (empty additional args to ndkstubgen)
2. module-libapi stubs that are also an ndk library (--systemapi --apex)
3. module-libapi stubs that are not an ndk library (--systemapi --apex
--no-ndk)
ndk_libs.bzl was used to differentiate between (2) and (3). This creates
an additional layer of indirection - users will need to modify this
external .bzl file if they would like to add a library to an ndk.
Replace this with an explicit atttibute in cc_stub_suite macro for better UX.
Test: go test ./bp2build
Test: b test //build/bazel/rules/cc:cc_stub_library_tests (added in
sibling CL)
Bug: 299501496
Change-Id: Idd3579e8013bae7a1740534f90d2767df5bac1a5
GLOBAL_THINLTO is now default, since it improves build speed, produces
more performant code and smaller binary. Remove the option to turn it
off since we no longer maintain that variation, and that simplifies the
logic (esp. for moving CFI builds to ThinLTO).
Also fixed a bug where ThinLTO is not propagated to its static deps on
non-default targets (32-bit or host).
Test: presubmit
Bug: 169004486
Change-Id: I31f41ba27c2b94a384d2ba5027049c307d6f4334
TARGET_VNDK_USE_CORE_VARIANT enables vendor to use some of the VNDK
libraries with core variant installed in /system/lib. However, this does
not make sense when VNDK is deprecated. This change is to ignore
TARGET_VNDK_USE_CORE_VARIANT when the VNDK is deprecated.
Bug: 303754049
Test: aosp_cf_x86_go_phone boot succeeded
Change-Id: Ie9fa75e0fa452e48924d51d64201690ffb271f33
The data attributes for cc library rules is dropped by their starlark
macros, so serves no real purpose. Furthermore, this unused dependency
proves problematic for allowlist v2, as there are many cases at HEAD
where the corner-case "requires" dependency would otherwise mark a module
as unconvertible.
Fixes: 303307456
Test: New unit test
Test: Manual verification to unblock allowlist v2 `bp2build.sh` runs in
AOSP
Change-Id: I6ca6104b958d2b428fc2ca5b3fa794106571acca
abort() seems friendlier, but in practice it puts a couple of irrelevant
libc.so frames on the top of the stack trace and clobbers all the
registers that might have helped remotely debug a sanitizer failure.
Let's switch to trapping directly in the sanitized code instead.
Bug: http://b/298741930
Test: treehugger
Change-Id: I6c971920c8506f97133749aaddd73b8cea6a12bc
This is now the default, and we will be removing non-GlobalThinLTO configuration.
Also changed handling for lto: never, it doesn't need to create an
lto-none variant.
Change-Id: I971baa920df867fb762923d925eed22215a89c27
Test: presubmit
Just knowing it failed is pretty useless. Tell people how to update
the ABI dumps.
Bug: http://b/156513478#comment43
Test: m ndk # after breaking the ABI
Change-Id: I7ed6a0acf2d1fbd942410678fc6aba2afbc5a2b6
This is the bulk of the "allowlist v2" feature. It will disable bp2build
generation for modules which have transitive dependencies without a
bazel build definition.
This CL includes this mutator, but doesn't register it as a bp2build
mutator (outside of a few unit tests). This allows us to easily iterate
on completion of this feature and ensure there are no launch blockers
before we finalize the change in AOSP.
Bug: 285631638
Test: Unit tests
Change-Id: Ifb0a079c409ca19b02cafa3fab2efa0d3deebc50
This module type is used by a single soong module - `common_libc` in
bionic/libc
Implementation details
- Convert this module type to ndk_headers rule. Bazel's ndk_headers rule
will have a boolean attribute `run_versioner` to determine if
verioner should be run on the headers
- Add this module type to the alwaysConvert bp2build list
- Add the converted target for `common_libc` to the deps of
`ndk_sysroot`. This ensures that unbundled apps link against the
versioned NDK headers of libc
Test: go test ./bp2build
Test: b build //bionic/libc:common_libc --config=android
Test: for f in $(find bazel-bin/bionic/libc/common_libc.versioned -type f); do cmp $f ${f/bazel-bin\/bionic\/libc\/common_libc.versioned/out\/soong\/ndk\/sysroot\/usr\/include}; done # no diff
Bug: 301169067
Change-Id: I55be202f0589db9bdc743c8be41c9c5afd74c352
PRODUCT_PRODUCT_VNDK_VERSION is set to 'current' by default. Now, we
can generate product variants without checking the
PRODUCT_PRODUCT_VNDK_VERSION build variable. Remove reading the
PRODUCT_PRODUCT_VNDK_VERSION variable from soong and generate product
variants by default.
Bug: 302255959
Test: m
Change-Id: I9a9b2076f4367c5ce9a393bbb206f8dee3884bd8
Test: go test soong tests
Test: verify no NeuralNetworksTest_random_graph sdk variants in
build.ninja when aosp_x86 is lunched
Bug: 302093869
Change-Id: I6429aa75518cca6974abcf7330236a6a07135ea1
Error message in case of not valid "nextAbiDump" included wrong
"prebuiltAbiDump.InvalidReason()". Refactor error message to include
"nextAbiDump.InvalidReason()" instead.
Change-Id: I2581734dd16649af2b282b3459f461658abd3d87
Signed-off-by: Aleksei Vetrov <vvvvvv@google.com>
This target will have a dependency edge to every bp2build equivalent of
Soong's ndk_headers. In b builds, sdk variants will compile against this
aggregated CcInfo providing target
A non monolithic alternative was discarded after conversations in
b/300504837#comment1-5
Contents of bp2build generated target: https://paste.googleplex.com/6643820291686400
Implementation details
- Since there is no equivalent Soong module for ndk_sysroot, hardcode
bp2build/build_conversion.go to collect all ndk_headers soong modules.
Add them to `deps` of a ndk_sysroot target
- Create `ndk_sysroot` in build/bazel/rules/cc/BUILD.bazel. This is
expected to be a temporary location. This will use the
cc_library_headers macro
- Update SetStubsForDynamicDeps so that sdk variant of rdeps depends on
//build/bazel/rules/cc:ndk_sysroot. This will provide a CcInfo during
compilation. Since ndk_sysroot is of type cc_library_headers, it will
not get packaged into the apk.
- Refactor `goBazelTarget` to a generic `bTarget` so that it is
representative of the expanded usage by ndk_sysroot
Test: b build //build/bazel/examples/android_app/java/com/app:app_with_sdk_variant_of_jni_deps --config=android (with aosp/2755284)
Bug: 300504837
Change-Id: Ifa427dd78115703ab251b0e1a0b71d3f19e91008
cfi and hwasan cannot be enabled at the same time. Skip building
vendor snasphots for the snapshot variants with both cfi and hwasan.
Bug: 277174974
Bug: 290779821
Test: HWASAN_INCLUDE_PATHS=frameworks/native/libs/gui m
Change-Id: Iaae913367a70e150d77a614de5f942aea98d1086
This should be no-op, as the underlying mutator has not changed yet.
Some other refactoring is required and done in this CL:
- Delete some old, dead ApiBp2build code
- Fix casting to TopDownMutator when it's not necessary
This change is required to prepare for allowlist v2 work, as only
BottomUp mutators can AddDependency.
Bug: 285631638
Test: m nothing
Test: presubmits
Change-Id: I5212a5f5634cc13056195783e6df37ff8eb000da
This is a followup fix for aosp/2753352 that added the name of the
`cc_stub_suite` as deps of `unbundled_app` config setting. However,
cc_stub_suite is really a macro that expands to <name>-<ver>.
Add the version explicitly to rdeps
Bug: 298085502
Test: b build //build/bazel/examples/android_app/java/com/app:app_with_sdk_variant_of_jni_deps --config=android (with aosp/2755284)
Change-Id: I8cf2804141bb14589e11b03a138b715538a3f1ee
This no-op refactoring facilitates some upcoming functional changes for
"bp2build allowlist v2". The work requires that the bp2build conversion
mutator be changed from a TopDown mutator to a BottomUp mutator.
Refactoring all bp2build-related methods so that they use Bp2buildMutatorContext
makes it easier to make this functional change without touching tens of
files and multiple projects.
Bug: 285631638
Test: m bp2build
Change-Id: I3d1ef3064146e959c6f0dc315350fc9764bf2bd2
Instead, we return an error. This allows us to access some product
variable information earlier when it will not be used as an attribute
without panicing
Test: m nothing
Change-Id: Id094b2b9e1364a8d174d99b3824fa149fb235b3e
Test: go tests
Test: CI
Test: patch in CL using prebuilts for mainline modules and build
Bug: 300640274
Change-Id: I73f70dac679e341670e1484ce57bc984f413d409
NDK has two ABI checks:
1. Check that prebuilt ABI exactly matches compiled binaries.
2. Check that ABI of level X compatible with level X+1.
The second check used compiled binaries for level X and prebuilts for
level X+1. But this approach may confuse people, because ABI change will
be shown "reversed". For example, adding new field to ABI monitored
binary of level X would be shown as "removed" when compared to level X+1
prebuilt.
This change uses prebuilts for both level X and level X+1. Given, that
prebuilts are checked to match compiled binaries, this should be enough.
Change-Id: If942e4319744bc5a2674cdd65f6a6f664fdfaa7e
Signed-off-by: Aleksei Vetrov <vvvvvv@google.com>
This reverts commit 49dcf79550.
The previous init crash was caused by the kernel defaulting to Sv57 and
the CFI shadow implementation in bionic assuming a 48-bit virtual
address space. With 5-level paging disabled in cuttlefish, we can again
enable CFI.
Bug: 293884796
Link: https://github.com/google/android-riscv64/issues/45
Test: cuttlefish boots
Change-Id: Ibcc02cf7ca6e396fd58674c648abf9f84daa260e
If a dependency has ndk stubs, then the sdk variant of the library
should link against the ndk stub variant and not the impl/apex_stubs variant
Unlike module-libapi, the depdendency does not go through an
@api_surfaces external repo indirection. This indirection was created to
support Multi-tree, and will likely be removed in the future
Test: Added a unit test
Bug: 298085502
Change-Id: Ie081e153fa586b6c22db0b8e42f91149fd8e5d9b
ndk_libary will be converted to a cc_stub_suite target. Its api_surface
attribute will be publicapi
The headers corresponding to this stub target will be added in a followup bug
(tracked in b/300504837)
Bug: 298085502
Test: Added a unit test
Change-Id: If9745083b18e0bcf5ecb89229a0f709b949d401c
This commit adds support for compiling rust rules inside the sbox
sandbox. To compile a rust module with sandboxing enabled, the entry
point to the crate must be specified via the `crate_root` property, and
all input sources and compile-time data must be specified via the `srcs`
and `compile_data` properties.
Bug: 286077158
Change-Id: I8c9dc5cf7578037a583b4be2e2f73cf20ffd4408
Ignore-AOSP-First: ag/24746588, in the same topic, is in an internal repo. This CL will be cherry-picked to AOSP afterward.
Test: go test
Bug: 277091218
Change-Id: I389d9535ea176991a1faa9beb46352b93363acd2
Merged-In: I389d9535ea176991a1faa9beb46352b93363acd2
cc_stub_suite now has an additional attribute correponding to the api
surface. This attribute will be used to pass additional args to the stub
generator.
Update bp2build to set the api_surface of the apex stubs. This ensures
that its stubs are generated with "--systemapi --apex"
Test: unit tests
Bug: 298085502
Change-Id: If4c479f85e6e485b5c795a565a0c559f1f013bf5
This module type will be converted to a cc_prebuilt_* bazel target. It
will provide
1. a prebuilt (.a/.so) file
2. headers (as -isystem)
Test: added a bp2build unit test
Bug: 298088835
Change-Id: Ib58cc7f6fde8f4ca34516f6f18a4c048a02a049a
Currently, we have ~4 of these modules which are all defined in
prebuilts/ndk. However, using ctx.ModuleDir() instead has the following
advantages
- makes bp2build simpler since we do not need to relativize this path
- prevents soong modules from reaching into another directory (as the
test setups were doing).
Test: m nothing
Change-Id: I780e2564cb37ebf4b800f0cd184789f3fc6f2fc8
Collect all NDK exported headers paths into a file that is used to
detect public types that should be ABI monitored.
Assume that we have the following code in exported header:
typedef struct Context Context;
typedef struct Output {
...
} Output;
void DoSomething(Context* ctx, Output* output);
If none of public headers exported to end-users contain definition of
"struct Context", then "struct Context" layout and members shouldn't be
monitored. However we use DWARF information from a real library, which
may have access to the definition of "string Context" from
implementation headers, and it will leak to ABI.
STG tool doesn't access source and header files, only DWARF information
from compiled library. And the DWARF contains file name where a type is
defined. So we need a rule to build a list of paths to public headers,
so STG can distinguish private types from public and do not monitor
private types that are not accessible to library users.
Bug: 156513478
Test: development/tools/ndk/update_ndk_abi.sh with enabled canDumpAbi
Change-Id: I9fa41e73450a41379638debb3dc56f421e0fb870
Signed-off-by: Aleksei Vetrov <vvvvvv@google.com>
For a dependency edge A --> B (stublib), Soong will link A's platform
variant against impl of B's platform variant if either of these are true
1. A and B have same apex_available
2. B has bootstrap: true
3. B is only available to platform
(3) was missing from bp2build. This CL adds that.
To implement this, we check the `apex_available` property of B.
Test: updated bp2build tests
Test: Built the internal module b/299191635 that was failing due to this
Bug: 299191635
Change-Id: Iafb173a3ab20d69b89f7949ce40c6f4096396f24
Soong currently adds -isystem
prebuilts/ndk/current/sources/cxx-stl/system/include to modules that
have ndk_system STL. This does not translate well to Bazel because of
its stricter sandboxing constraints.
In preparation for building sdk variants with Bazel, create an
indirection where sdk variants that use this STL depend on a
`ndk_system` header library module.
This should be a no-op in Soong, but the resultant ninja files are not
identical. (e.g diff https://diff.googleplex.com/#key=OOLtc1GFmDDF). The
-isystem now appears _before_ the local cflags. However, this should be
fine because we have a check in `CheckBadCompilerFlags` that bans use of
-I/-isystem in user-provided cflags.
Test: diff'd ninja file
Test: TH
Bug: 298258442
Change-Id: I8c4e2b66bb9ac25c44ceedd52298ba474a554a04
This enables Machine Learning Guided Optimizations (MLGO) for register
allocation. App launch performance on oriole-userdebug is improved by
0.53%, and binary size is very slightly reduced (e.g. 64 bytes for
libart.so).
https://dashboards.corp.google.com/_0e1b06dd_a79e_4433_b625_b62e5cac004b?p=BootImageProfileId:36265
Test: presubmit
Bug: 229056536
Change-Id: I0e5d40bb90a1555a1343f114ebffffca281f20b2
-Wmultichar is enabled by default, so we don't need to explicitly
enable it.
Bug: 298703966
Test: presubmit
Change-Id: Iae820145e6412365024a97ab3a34d760563785fa
