Move the ImageMutator to be registered just after the archMutator
in preparation for moving it between osMutator and archMutator.
Requries updating variants in a few tests that now run the
ImageMutator.
Bug: 142286466
Test: no change to build.ninja
Test: all soong tests
Change-Id: Ia9d2a7bc0e225bedec3c9a83ea04f471a931bf47
This function took a props varargs parameter but it was only ever
called with no parameters. The props parameter meant it could not be
used directly with RegisterModuleType() so the defaultsFactory() func
was added which simply called through to DefaultsFactory().
Removing the props varargs parameter allowed it to be used directly
with RegisterModuleType() and so the defaultsFactory() could be
removed.
This was needed because a follow up change adds a test in another
package that uses java_defaults and so it needs the factory method
(defaultsFactory()) to be exported.
Test: m nothing
Change-Id: I33d5c4d4ce8f349b1e6fc706d5e9656faf303b4f
Add a new java_header_libs property that exports header jars (as
java_libs currently does) and switch java_libs to export implementation
jars instead.
Refactors implementation of the existing library sdk member type so
both properties can be supported from common code as they differ only
on the jar being exported.
Bug: 143678475
Test: m nothing
Change-Id: I04642122f72d083bbdfd3290624f957b71ee8875
* changes:
Organize sdk member properties
Extract the cc and java sdk related tests out into their own file
Improve testing of sdk snapshot generation
Separate sdk testing infrastructure from sdk tests
Parameterize the sdk member processing
Link type checking for java_library has not been working unintentionally.
So turn on link type checking for these types.
And also add tests for link type checking.
Bug: 145799020
Test: cherry-pick aosp/1182522 and check if build fails
Test: m nothing and there is no error(soong unittest)
Change-Id: Ifc347f657885de1028ac0076ddd103c0387b597a
This was broken after CL:
Ida40dfae8c83bf7c2e737d5c7ea418e1197ad826
..which split boot image into primary and extension.
The argument to oatdump was not updated properly.
The old form before the split was:
--image=out/soong/generic_x86/dex_bootjars/system/framework/boot.art
The new form after the split is:
--image=out/soong/generic_x86/dex_artjars/apex/com.android.art/javalib/boot.art:out/soong/generic_x86/dex_bootjars/system/framework/boot-framework.art
Test: lunch aosp_x86-userdebug && m dump-oat
Test: Manually inspect build command to make sure correct image
location is used:
$ fgrep -e 'oatdump' $ANDROID_BUILD_TOP/out/soong/build.ninja | egrep -e '--image[^ ]*' -o
--image=out/soong/generic_x86/dex_artjars/apex/com.android.art/javalib/boot.art:out/soong/generic_x86/dex_bootjars/system/framework/boot-framework.art
Change-Id: I54a5146c404aefb6eaeadbe3e9363db935ecad86
Grouping in alphabetical order by package and then by name within the
package should minimize conflicts when making changes.
Bug: 143678475
Test: m conscrypt-module-sdk
Change-Id: Ia7dbcd41ce8b8dd8675a90b1b6868fcaeaf72ee4
Makes it easier for multiple people to work on different areas without
conflicts.
Bug: 143678475
Test: m conscrypt-module-sdk
Change-Id: I93140450c0b049946a9d0432225bccff82f5a873
Adds TestHelper to provide general test helper functionality for use by
any test.
Adds testSdkResult, composed with TestHelper that encapsulates the
result of processing and sdk {..} definition and provides specialized
support for testing the build rules.
Dedups the analysis of the sdk build rules, and improves it to extract
more information, and in different forms. That is represented by the
snapshotBuildInfo struct.
Adds a CheckSnapshot() method which checks the snapshot for a specified
sdk version. It takes a list of functions that can each perform a check
on a specific facet of the supplied snapshotBuildInfo.
Methods are provided for tests to use to check the following facets:
* Generated Android.bp contents.
* Copy rules
* Merge zip inputs
This approach makes it possible for each test to customize what is
being checked without either duplicating functionality, causing a
proliferation of specialized forms of the CheckSnapshot method for
different types of tests or adding arguments for every possible check
that any test would need which would lead to lots of churn to existing
tests when new arguments are added.
The main testing improvement is for CheckSnapshot() to actually try and
load the Android.bp that is generated. In order to do that it was
necessary to create a mock filesystem populated with information from
the build rules, i.e. the destination files from every Cp command as
well as the destination directory from every repackage zip command.
That helps detect a number of sources of errors:
* Failing to copy a file/directory that is mentioned in the generated
Android.bp file.
* Invalid properties.
* Invalid format of the .bp file.
* Integrity issues within the .bp file.
Bug: 143678475
Test: m conscrypt-module-sdk
Change-Id: I4d3fe18f86698186d18e7e8b32d2e319183f7f0c
Extracts the type specific functionality into the SdkMemberType
interface which has to be implemented by each module type that can
be added as a member of the sdk. It provides functionality to add
the required dependencies for the module type, check to see if a
resolved module is the correct instance and build the snapshot.
The latter was previously part of SdkAware but was moved because
it has to be able to process multiple SdkAware variants so delegating
it to a single instance did not make sense.
The custom code for handling each member type specific property,
e.g. java_libs, has been replaced with common code that processes
a list of sdkMemberListProperty struct which associates the
property (name and getter) with the SdkMemberType and a special
DependencyTag which is passed to the SdkMemberType when it has to add
dependencies.
The DependencyTag contains a reference to the appropriate
sdkMemberListProperty which allows the resolved dependencies to be
grouped by type.
Previously, the dependency collection methods would ignore a module if
it was an unsupported type because they did not have a way of
determining which property it was initially listed in. That meant it
was possible to add say a droidstubs module to the java_libs property
(and because they had the same variants) it would work as if it was
added to the stubs_sources property. Or alternatively, a module of an
unsupported type could be added to any property and it would just be
ignored.
However, the DependencyTag provides information about which property
a resolved module was referenced in and so it can detect when the
resolved module is of the wrong type and report an error. That check
identified a bug in one of the tests where the sdk referenced a
java_import module (which is not allowed in an sdk) instead of a
java_library module (which is allowed). That test was fixed as part
of this.
A list of sdkMemberListProperty structs defines the member properties
supported by the sdk and are processed in order to ensure consistent
behaviour.
The resolved dependencies are grouped by type and each group is then
processed in defined order. Within each type dependencies are grouped
by name and encapsulated behind an SdkMember interface which includes
the name and the list of variants.
The Droidstubs and java.Library types can only support one variant and
will fail if given more.
The processing for the native_shared_libs property has been moved into
the cc/library.go file so the sdk package code should now have no type
specific information in it apart from what is if the list of
sdkMemberListProperty structs.
Bug: 143678475
Test: m conscrypt-module-sdk
Change-Id: I10203594d33dbf53441f655aff124f9ab3538d87
It is easier to extract information out of static build rules than it
is out of custom build rules built using the builder as the former
provides access to the in/out and args separate from the rule whereas
the latter only provides access to in/out.
Also, cleans up some warnings that appear in Intellij.
There is a lot of duplication in the testing code. That will be
resolved in a follow up change.
Bug: 143678475
Test: m conscrypt-module-sdk
Change-Id: I973bc0c90b0affd84487f1b222dd3e6c22c07ec0
This is so that we can split the build_test build on large branches into
multiple builds, each testing a subset of the products.
Test: m blueprint_tools (runs the go unit tests)
Test: build/soong/build_test.bash --only-config
vs
Test: build/soong/build_test.bash --only-config --shard-count=2
Test: build/soong/build_test.bash --only-config --shard-count=2 --shard=2
Change-Id: I40ccc1aa477bc0ffa74ff564d155068509be18f0
Add a cuj_tests binary that runs through a sequence of critical
user journey builds and collects metrics.
Bug: 145688720
Test: cuj_tests
Change-Id: I6bb1a3fc4a458fd8cea7b7c924f565d0cf3e6df3
cc_fuzz ignored soong namespaces, and built all cc_fuzz modules into the
architecture package. This caused naming conflicts where two modules had
the same name but were in different namespaces.
Now, cc_fuzz will only build modules that are in the same namespace as
the product definition - resolving the duplicate definitions error.
Bug: 145240569
Test: m haiku
Change-Id: I0a0acd139426301586ac5e3aba8ce5527cb7a1b5
This change fixes the problem that when an apex module is overridden by
another override_apex, the <apex_name>-file_contexts are duplicated when
creating the system-level file-contexts.
Fixing this by not emitting the file_context info for the overridden
apex.
In doing so, OverridableModule interface was extended to have
GetOverriddenBy() method which can be used to test whether a module is
an overridden one or not.
Bug: 144338929
Test: m (apex_test amended)
Test: add "override_apex {name:"com.googlge.android.tzdata",
Change-Id: I5e9401c32899bb9987c90cba4185f571dc1a87f0
base:"com.android.tzdata"}" and the build is successful
Exclude vendor/ and hardware/ from the manually written binder interface
whitelist.
Bug: 136279235
Test: interfaces in excluded directories are not checked against whitelist.
Change-Id: I0d640e23489b37d0c0787d5fca6bcdab10034109
The use of pragma pack(1) affects the packed-ness of every struct
definition that follows it in the translation unit, even in following
header files. This can lead to multiple translation units disagreeing
on whether a struct is packed.
Enable the warnings (as errors) that help spot this problem.
Bug: http://b/73971777
Test: treehugger
Change-Id: Id3096f35dff5dc8723d30d1305f1ed7ae8190a32
Building sdk snapshots on anything other than linux glibc is not
supported at the moment.
Bug: 145598135
Test: m nothing
Change-Id: I34155252545e9f37c5361ed94a27a87024fcece2
When flattened, the Make module for an APEX is a phony package for the
files in it. There is no output (either implcit or explicit) there.
Bug: 144338929
Test: m
Test: check the generated Android-<target>.mk file
Change-Id: I86df3b69f402a7e2d94bafa39f2aad3312d8c28b
This patch splits the system boot image in two parts:
- The ART boot image. This is the primary boot image that is
included in the ART apex and contains dexpreopted Core Libraries.
- The framweork boot image extension. It depends on the ART boot
image and contains framework libraries.
The third "apex" boot image (used in the JIT-zygote experiment)
remains unchanged; it is a monolithic primary boot image that
contains both libcore and framework libraries.
Dexpreopting of APKs now uses the framework boot image extension
(which in turn pulls in the ART boot image as a dependency).
Test: m
Test: phone boots:
lunch aosp_walleye-userdebug && m \
&& adb reboot bootloader && fastboot flashall -w
Bug: b/119800099
Exempt-From-Owner-Approval: rebased after getting approval.
Change-Id: Ida40dfae8c83bf7c2e737d5c7ea418e1197ad826
Sasha Smundak