Currently, there is little verification around allowlisted modules
actually being mixed-built. This flag would allow us to verify
that a module allowlisted is mixed-built for at least one variant.
Bug: 278910100
Test: m nothing --bazel-mode-staging --ensure-allowlist-integrity
Test: m nothing --bazel-mode-staging --ensure-allowlist-integrity
--bazel-force-enabled-modules=com.google.android.neuralnetworks (This
fails, as expected)
Test: build/soong/test/mixed_mode_test.sh
Change-Id: Icd5976f4f44f1a8caca1e5247d986642f7995f97
This was disabled because of some non-determinism in Soong, which has
been been fixed in aosp/2559590
Bug: 275313114
Test: go test ./cc -run TestMixedBuildUsesStubs -count 1000
Change-Id: I261dc785c832e66f8e213132c19111b8b90e40e9
In addition to the license text files, include license kinds
information to json file. Also, use the original paths of license
text files that are copied to NOTICE_FILES directory.
This will be used when generating snapshots from the snapshot build.
Bug: 271539873
Test: m vendor-snapshot
Change-Id: I0c0427bb66f2c1fca322d5554aa66220a3b62fb3
This reverts commit 1db4348734.
Changes from original:
- extracted function to add the current version to stub
versions for bp2build.
- added libc++ to mixed builds denylist
Reason for revert: re-uploading with fix in topic
Change-Id: Ifa0ed456bf8cb4a7f861d6826263adfedb4fdd9c
Modifying the list in-place causes some non-determinism.
Test: go test ./cc -run TestStubsForLibraryInMultipleApexes -count 1000
Bug: 275313114
Change-Id: Ia2519e146d97667ad5900cf68ab9935fcbaf08a4
The Bazel rules don't currently support any sanitizers other than ubsan,
so we should disable mixed builds for modules which are sanitized.
Test: go test
Bug: 278772861
Bug: 253433725
Change-Id: Ia01fb8cb59154bdfb21a111b04af0350e1876b0b
system libraries (libc, libm, libdl) belong to the runtime apex api
domain. platform and other apexes should link against stubs and not impl
of these libraries. Update bp2build so that both apex/non_apex picks
stubs. These axes need to be more granular in the future, but for now
this works because
- non_apex axis (platform) is a separate api domain than the runtime apex
- com.android.runtime which matches apex axis has not been allowlisted
yet.
Test: TH
Change-Id: I43f429a90875f30d176db17b144dd51674ec8adf
We used to enable unique names only when a profile is available.
But A profile generated for a binary without unique names doesn't
work well when building a binary with unique names. To avoid a
chicken-and-egg problem, this CL enables the flag when afdo=true,
whether a profile exists or not.
Bug: 241523910
Test: build
Change-Id: I74d834510d81d2db76e34d0488c74a60e1fcecd5
Previously we were treating it as a shared library only, but
cc_test_library produces both a static and shared library.
Test: bp2build tests
Change-Id: I293d4246d581212b95765fa8f084301514d00dbf
in bp2build. The context here is that `-fvisibility=default` should
only be added for CFI if `-fvisibility=hidden` is not already
specified. This will be achieved using toolchain features. Note
that Soong itself never adds `-fvisibility=hidden`. This is only
ever added in the `cflags` property of a bp file.
Bug: 261733820
Test: Unit tests
Change-Id: Ib821e8c30a9cd03d2929b4bd2e771bec7b33fa66
We can just rely on the prod/staging allowlists which name the modules directly.
Test: presubmits
Test: m nothing; showcommands adbd_test | grep bazel-out
Fixes: 279004435
Change-Id: I666855d6bbe63c628256c3eebd7ae581c0cbdedb
Environment variables HWASAN_INCLUDE_PATHS and
PRODUCT_HWASAN_INCLUDE_PATHS can be used to enable HWASan for multiple
modules, by just adding the module directory to the env variable.
Bug: b/271948407
Test: Set specific module directory to above env variable and check the
assembly codes of output elf files after building, finding hwasan
related symbols inside.
Change-Id: Ic49b515830c4469ca5fa94f547b26c0fb602fc54
This reverts commit ca438e6b72.
Reason for revert: ci post-submit failures in aosp-master-bazel of targets mixed-droid-clean and mixed-droid-incremental:
FAILED: ninja: 'out/target/product/generic_arm64/obj/SHARED_LIBRARIES/libc_intermediates/libc.so.toc', needed by 'out/target/product/generic_arm64/obj/EXECUTABLES/updater_intermediates/LINKED/updater', missing and no known rule to make it
12:21:27 ninja failed with: exit status 1
Change-Id: I081b499d23f2568cdf6227c4e3b0278164086b69
ld -r reorders symbols and invalidates the .llvm_addrsig section, which
then causes warnings if the resulting object is used with ld --icf=safe.
The warning is especially common when building with musl, as the
clang_rt.crt* objects have .llvm_addrsig sections, are linked into
libc_musl_crt* using ld -r, and are then linked into every other binary
and shared library with --icf=safe.
Strip the .llvm_addrsig section after ld -r to prevent the warnings.
Test: m USE_HOST_MUSL=true host-native -k
Change-Id: Ia52a4756b9ebbb62115898d0de9f8641e6fea705
Revert submission 2478283-master-I3cf18e7814ff52d526017e4fe34b17ab3d8f1080
Reason for revert: implementing a different way
Reverted changes: /q/submissionid:2478283-master-I3cf18e7814ff52d526017e4fe34b17ab3d8f1080
Change-Id: I1d690b6d322adc664e4c02d958f0183bbb78e669
Bug: 270408757
Test: go test
Test: m NeuralNetworksTest_shared_partial --bazel-mode-staging
&& verify that Ninja commands link libneuralnetworks via stubs
Change-Id: I19bf5a829cea7fd00c0f82511ad2bddbc40fae4f
The public stubs of libclang_rt.* originate from the same module. When
generating the Androidmk entries for these stubs, we need to add a
suffix to differentiate between the two stub variants.
This is likely an issue specific to libclang_rt.*. The public stubs of
other libraries like libc, liblog, ... come from a separate module
ndk_library which adds an .ndk suffix
Test: lunch module_arm64only && TARGET_BUILD_APPS="build.bazel.examples.apex.minimal" build/soong/soong_ui.bash --make-mode nothing
(with aosp/2535141)
Test: lunch aosp_cf_arm64_phone-userdebug && m nothing
Change-Id: I2b4411cb022b4e27f44d2fce6cb26492dd7e2278
We're switching from x18 to gp, in stages. This change frees up x18
again now bionic and ART have switched to gp. There's no need for a
corresponding "reserve gp" change because gp is automatically reserved.
That's what makes it the perfect choice for shadow call stack (at
least until we get Zsslpcfi and an actual hardware shadow stack pointer
special-purpose register).
Bug: https://github.com/google/android-riscv64/issues/72
Bug: http://b/277909695
Test: treehugger
Change-Id: Iaee1d44cba8daf5d4a27a3791fe3cc28ed1a5d9a
We're trying to switch from x18 to gp for shadow call stack, ideally
without breaking anything, so let's disable shadow call stack while we
rearrange things behind the scenes...
Bug: http://b/277909695
Test: treehugger
Change-Id: Ic1a6bceaf0ab20dc66e1903747ebb668232a253a
This feature has a substantial impact on RSS memory usage (>2% in many processes). Turn this off for now.
Test: RSS memory benchmark
Bug: 277565884
Bug: 277083500
Change-Id: I2fc5826048ba5d00cabcc8401ac40a6f54812a43
VNDK snapshot must include the information of the license texts and
kinds. Include them when packaging the snapshot prebuilt files.
This will be used to generate the VNDK snapshot files.
Bug: 270625053
Bug: 277317599
Test: development/vndk/snapshot/build.sh --build-artifacts
Change-Id: I0a39ff598efbb80faaf63807cdf5ce685fa074cc
libc_hwasan is a new library in the runtime apex that lives in
bionic/hwasan/libc.so and is symlinked to /system/lib64/hwasan/libc.so.
This is chosen by the linker if an app or binary requires HWASan
support.
Bug: 276930343
Change-Id: If331744ad84241ad99a41805ea3110d37cf9b0af
This enables turnning off -fprofile-sample-accurate on a per project
basis.
Bug: 274725934
Bug: 273415627
Test: presubmit
Change-Id: I15060950f7b55e51069cb00b66512c7e94981f1b
