The wrapper to getenv() appends the effective userid (euid) of the
current process to GCOV_PREFIX. This avoids conflicts and permissions
issues when multiple processes try to create/access the same directories
and files under /data/misc/trace.
Test: Verify that coverage files are written to
/data/misc/trace/<euid>/proc/... instead of /data/misc/trace/proc/...
Change-Id: If58081a12b2b9bc40cfcbf64c99beafe198d07af
Bug: http://b/134177005
Bug: http://b/116873221
Previously, the libprofile-extras dependency was added as a
LateStaticLib and the constructor in this library was included during
linking with the '-uinit_profile_extras' linker flag. This was done
because at the deps() stage, the exact binaries that need coverage are
not known (in fact the coverage-enabled variants are not created yet).
This meant that for a link command, if one of the shared libraries
already exported the constructor, the output of the link command did not
load/link libprofile-extras.
For other reasons, we now want to add more symbols to this library that
need to be linked into all libraries and executables. To accomplish
that, refactor the dependency handling so libprofile-extras can be added
as a 'WholeStaticLib'.
This is done by creating a new dependency type (with a coverageDepTag
dependency tag) to add libprofile-extras as a dependency for all modules
that can potentially link with coverage. During the flags() call, this
dependency is moved as a WholeStaticLib dependency iff coverage is
enabled in this link step.
There are a few NFC changes as well:
- deps() takes a DepsContext parameter.
- flags() has an extra PathDeps parameter and return value.
- add useSdk() helper to cc.Module.
Test: Build with coverage and check that we can generate coverage using
SIGUSR1 and the debug.coverage.flush sysprop.
Change-Id: I7e7d8201956a150febbda5bb1794f8ece016db8b
Library components can enable the flags based on
performance impact
Test: run-test gtest
Change-Id: I41c4464cffd2b0e830d4ab1c5a25b378e2a25129
Signed-off-by: jaishank <jaishankar.rajendran@intel.com>
Bison is currently looking for m4 directly in `/usr/bin/m4`, instead of
through $PATH, so it hadn't been tracked via the $PATH interposer
checks. But if you don't install m4 at all, it starts failing.
Bug: 117561006
Bug: 130111713
Test: Run on a machine without m4, build passes
Change-Id: I3fbad64127c183de79c17511e979fd5df946eb80
This commit enables ABI diff for libraries without source.
Stub libraries are excluded.
Test: ./create_reference_dumps.py -products aosp_arm64
Test: make
Bug: 135728451
Change-Id: Idbe276db6f5e057175fc098f0226cb38996faa17
sanitizerRuntimeDepsMutator only modifies the currently visited
module, it can visit modules in parallel.
Also, stop recursing into modules that are not static dependencies,
and stop recursing if the module already has all modifications that
the mutator could make.
Test: m checkbuild
Change-Id: I95a57f763a91940f1854ba3c587a2f70e8baba97
Sanitizer coverage is currently broken with the experimental pass
manager. See b/133876586 for more information. The patch is currently
being worked on upstream (https://reviews.llvm.org/D62888), but is not
ready yet. Hence, we disable it here (similar to the LTO bug).
Bug: 133876586
Test: Build anything with SANITIZE_TARGET='fuzzer', verify that it has
'sancov' symbols.
Change-Id: I4302a8cf05300015aaff137f40b61d5685dc3fea
Current Android toolchain does not support hwasan with new pm.
Bug: 135298400
Test: manual, inspect the binaries for calls to __hwasan_init
Change-Id: I126c59e6ecabc3a1699e1bf34c546568b6e8a21f
require_root is added to allow auto-generated test config to include
RootTargetPreparer so the test runs with root permission.
Bug: 134509111
Test: add "require_root: true" to init_benchmarks and libpower_test
build the modules, confirm the extra target preparer is added in the
test configs.
Change-Id: Ia07503e338935d6aa92560e7cf7b18d2a4c51243
Instead of having all the object compilations depend on a stamp file,
instead depend on the headers, which are already available in the build
graph.
This allows my RBE builds to just find the headers, instead of having to
walk the graph to find the other outputs.
Test: treehugger
Test: build a system image with RBE
Change-Id: I0a310229be26d7bd399e074a4350d5be038f8661
* changes:
Consolidate *MutatorContext and ModuleContext into BaseModuleContext
Add test for missing defaults modules with AllowMissingDependencies
Capture missing dependency error rules
Share buildDir for android/soong/android tests
Reimplement AddMissingDependencies
Add GenerateAndroidBuildActions to DefaultsModuleBase
genrule modules were not mutated by the apex mutator. As a result, a
module that is mutated for an apex have depended on the genrule modules
host_bionic_linker_{asm|flags} that are not mutated for the APEX.
This in turn caused an inconsistency problem because the
host_bionic_linker_* genrules use the dynamic linker that isn't mutated
for APEX as an input, while the outputs of the genrules are used by
modules mutated for APEX. The inconsistency caused b/135008828.
Fixing this issue by correctly let genrule to be mutated by the apex
mutator.
Bug: 135008828
Test: checkout master-art-host
ALLOW_MISSING_DEPENDENCIES=true DIST_DIR=out/dist
./art/tools/dist_linux_bionic.sh -j50 showcommands
com.android.runtime.host
is passing
Change-Id: I92d90400c61d3951f6202d932832d0cb63c95a7d
In the first design, public sysprops have been accessible from Java
modules linking against SDK. But SDK modules shouldn't do, because
sysprop_library isn't for the apps. This renames system to public, so
that only public(System till now) and internal scopes remain from now.
Bug: 131637873
Test: m && sysprop_test
Change-Id: I548007d4a6018922f98d3d13915cee1d66070086
