Ignore overlay directories that have been selected for enforced RRO
by the product, and pass them to Make instead to be converted to
an auto generated RRO package.
Bug: 69917341
Test: m checkbuild
Change-Id: I8e2677f4c600acdd8dee0869bf4fbc3d5dbc8b44
The system_$(VER) is added for vendor, similar to sdk.
Bug: 67724799
Test: build
Merged-In: I2545c92707591ca278066870c74e9f49e9825855
Change-Id: I2545c92707591ca278066870c74e9f49e9825855
(cherry picked from commit b8baff1fa3)
Update app support enough to build framework-res.apk, link
framework.jar against its generated files, and export it to
make.
Bug: 69917341
Test: m checkbuild tests docs
Change-Id: I7db29cd1f5fabb22e844483ecc7c38abfedbbe0a
AConfig() now duplicates Config(). Replace the uses of AConfig()
with Config(). Leave AConfig() for now until code in other
projects is cleaned up.
Test: m checkbuild
Change-Id: Ic88be643049d21dba45dbd1a65588ed94bf43bdc
In Soong, a Config() method will always return a Config. Make
ModuleContext, SingletonContext, TopDownMutatorContext and
BottomUpMutatorContext's Config() methods explictly return
a Config to avoid having to type-assert everywhere. Overriding
the Config method requires duplicating the list of methods in
blueprint.BaseModuleContext and blueprint.BottomUpMutatorContext,
following the same pattern used by the other *Contexts.
Config() obsoletes the AConfig() method used in some places, which
will be cleaned up in the next patch.
Test: m checkbuild
Change-Id: Ibe21efde933959811d52443496967ab8ce71215e
This reverts commit 63a250a336.
Reason for revert: Some failures:
namespace_test.go:648: dir1/Blueprints:2:4: a namespace must be the first module in the file
such as New Build Breakage: aosp-master/build_test @ 4475274
Change-Id: I1b5db8eb934e51ff22241bfca44199d886b1393b
Bug: 65683273
Test: build/soong/scripts/diff_build_graphs.sh \
--products=aosp_arm \
'build/blueprint:work^ build/soong:work^' \
'build/blueprint:work build/soong:work'
# and see that the only changes were:
# 1. adding some new files
# 2. changing some line numbers
Test: m -j nothing # which runs unit tests
Change-Id: I6d3e52ef62c4cabe85b9a135a54de0e1a6aab29c
Use aapt2 instead of aapt to compile Android app resources.
Also generate all files into srcjars instead of individual
sources.
Test: m checkbuild
Change-Id: I5a67991a0daf0017e8159b46fcff7d5564a91468
Wrap blueprint.PackageContext so that the *Func methods can provide
an android.Config instead of an interface{}. The modified signatures
means that every method in ModuleContext and SingletonContext
that takes a blueprint.PackageContext now needs to be wrapped to
take an android.PackageContext.
SingletonContext wasn't previously wrapped at all, but as long
as it is, wrap everything like ModuleContext does. This requires
updating every Singleton to use the android-specific methods.
Test: builds, all Soong tests pass
Change-Id: I4f22085ebca7def6c5cde49e8210b59d994ba625
Bug: 64267858
Test: create reference abi dump for libjpeg locally. Change return type
of libjpeg_std_err and build with :
BOARD_VNDK_VERSION=current mm -j64 showcommands dist
Build fails and abidiff report gets copied into out/dist/abidiffs.
Change-Id: I7c8ecfac95361e731009e5913bd3a7bb323a9597
Can be caused by `json:"omitempty"` instead of `json:",omitempty"`
Bug: 69076024
Test: m -j # which runs unit tests
Change-Id: I92e3193d00a740c72d36a56748e0b0a8ad1d772e
* changes:
Fix genrules depending on Go tools
Use logtags without merged file
Export extract-srcjars.sh to Make
Remove partial javastream proto support
Add support for .srcjar files from genrules and srcs
Fix java AIDL properties to match C/C++
Fix proto flags in java
genrules lost the ability to depend on Go tools after
I05e945f38915d49cd3c0ab72a86576949bc7eff2 which converted
VisitDirectDeps from blueprint Modules to android Modules.
Add VisitDirectDepsBlueprint to visit all modules including
blueprint Modules, and use it in genrule. Also add a check
for disabled modules that was being handled by
VisitDirectDeps.
Test: m checkbuild
Change-Id: I65724283166c63596d071e598c08fed87ef32896
This fourth possible value currently has the same semantics
as a default/unset EXPERIMENTAL_USE_OPENJDK9, but allows
people to explicitly switch back to the old semantics when
the default changes.
Test: make showcommands core-oj (in environments with
EXPERIMENTAL_USE_OPENJDK9 set to "", "1.8", and "true").
Bug: 38177295
Change-Id: I25accf14344a05349a6e97572d7c2c1f6a7f2063
Add an android.DirectorySortedPaths that stores paths sorted such
that all paths in a directory including subdirectories are in a
contiguous subslice. This will allow efficient O(log(N)) finding
of all paths in a directory using a binary search on the directory
prefix.
Test: TestDirectorySortedPaths in paths_test.go
Change-Id: I5a06a89351ae06e88c06526be54a6b79075361b7
This isn't often used, but the values were quite old. So update them to
something closer to aosp_arm64 from today.
Bug: 68803744
Test: m clean; m --skip-make libc
Change-Id: Id5c60b62e7c7a91635dea9669fc604d63d299778
This CL adds the ability to centrally enable or disable CFI for
components using either an environment or product config
variable. This is a better, nore manageable option that enabling CFI
across each component individually.
Bug: 67507323
Test: CFI_INCLUDE_PATHS= system/nfc m -j40
Test: CFI_EXCLUDE_PATHS = frameworks/av m -j40
Change-Id: I38b77946759121aec99ab25f31aaee2e5d993f73
there's no use case for prepending/appending to bool, and string
properties within module struct. Declearing "*bool" and "*string" almost
cover everything user need.
I did see one case that user specify relative_install_path as
path prefix in cc_defaults, and concatenate with the one in real module
to get the final relative install path in Android.bp <bionic/tests/libs>.
Test: m -j checkbuild
Bug: b/68853585
Change-Id: If3a7a2689c3fc307aae136af6bc9c57f27a1e1a0
Allow functions to get the result associated with a OncePer key without
also specifiying a function. Panics if the key has not already been
set. Also replace the open-coded concurrent map implementation with
the new sync.Map.
Test: m checkbuild
Change-Id: I814fdb1ffffaee8398dc877af146e29638c8a6a8
Move AppendProperties and PrependProperties into TopDownMutatorContext
so that LoadHooks can be a subset of TopDownMutatorContext that also
includes CreateModule.
Test: m checkbuild
Bug: 35570956
Change-Id: Iffa6a6aec96f08821c2446e0e0f4622ab772b54c
This reverts commit 5c3c768187.
This reapplies I70e45caa92f65dc6a3bc81ae33a5a8f34ad11890 along with
additional changes in other projects to match the new types.
Bug: 68337468
Test: m checkbuild
Change-Id: I2d2d876b291869f834a6468d3e2ad99148f94a15
Strip debug info in javac when PRODUCT_MINIMIZE_JAVA_DEBUG_INFO
is set.
Test: m with PRODUCT_MINIMIZE_JAVA_DEBUG_INFO=true
Change-Id: I167e742662801291c516bf1ff826486560d22147
Switch commonProperties to *string and *bool so that defining them
in a module correctly overrides instead of appends to values set
in defaults modules.
Bug: 68337468
Test: m checkbuild
Change-Id: I70e45caa92f65dc6a3bc81ae33a5a8f34ad11890
A module referenced with srcs: ":module" was not honoring
ALLOW_MISSING_DEPENDENCIES=true. Don't fail if the module
doesn't exist, an error will already have been produced by
ExtractSourcesDeps.
Bug: 68183622
Test: m ALLOW_MISSING_DEPENDENCIES=true
Change-Id: Id806633d12ab2ecd78b532e1922e59824e5c20a8
Also adds checks that the dependencies are android.Modules and
are not disabled.
Test: m checkbuild
Change-Id: I05e945f38915d49cd3c0ab72a86576949bc7eff2
Move firstUniqueElements to android.FirstUniqueStrings,
lastUniqueElements to android.LastUniqueStrings, and lastUniquePaths
to android.LastUniquePaths.
Test: m checkbuild
Change-Id: Ieac840405126c7f8f98afb4a4ef35c01a18fe7fb
Now that android.ModuleContext does not include blueprint.ModuleContext
we can rename android.ModuleContext.ModuleBuild to
android.ModuleContext.Build without colliding with
blueprint.ModuleContext.Build. Leave ModuleBuild as a wrapper around
Build for now to avoid having to update all the users outside
build/soong simultaneously.
Test: m checkbuild
Change-Id: I18eb8cc04faf002049a11d9aac97e9732ff5d638
android.ModuleContext should replace the blueprint.ModuleContext
methods with ones that take android.Module parameters instead of
blueprint.Module parameters. As a first step, don't include
blueprint.ModuleContext in android.ModuleContext and manually
specify the methods from blueprint.ModuleContext that we need.
Future patches will replace these methods with ones that take
android.* types. Also do the same for TopDownMutatorContext.
Test: m checkbuild
Change-Id: If11c298eed98bc65712b4942d0283cafa058726c
The source for dx may not be available in PDK builds, use the
prebuilt one from prebuilts/build-tools instead.
Bug: 67663308
Test: m TARGET_BUILD_PDK=true
Change-Id: I9090b5190539f901fc05264a472133c12d4ea2a1
If sdk jars(android_stubs_current, etc) are compiled using soong java
modules, we have to filter them when running Java build with Turbine.
TODO: provide more unit-tests.
Test: m clean && m -j32; go test java_test
Change-Id: Iad7c241b0e8b0ca760950733f513124b56c84564
Make Rel() on ModuleOutPath and ModuleGenPath return the path
relative to the module out and module gen directories respectively,
and make TestingModule.Output() match against the full relative
path to the module.
Test: java_test.go still passes
Change-Id: Id5b2ec3fdef41d4169b943e68d032fc64a2b6f92
This uses knowledge of transitive dependencies to reorder
linker command line arguments such that if module A depends
on module B, then module A is automatically listed before
module B in the linker command line.
This should mostly remove the need for Android.bp files to
list all of their static dependencies in link order
Bug: 66260943
Test: reorder the entries of static_libs in an Android.bp and see that linking still succeeds
Change-Id: I20f851ab9f2f30031254e4f30023b6140d15d6c3
Allow java libraries to specify .kt sources, precompile them with
kotlin, and then pass them in the classpath to javac.
Bug: 65219535
Test: java_test.go
Change-Id: Ife22b6ef82ced9ec26a9e5392b2dadacbb16546f
Adds a java_system_modules module type that (when
EXPERIMENTAL_USE_OPENJDK9 is set to true) converts a list of
java library modules and prebuilt jars into system modules,
and plumbs the system modules through to the javac command
line.
Also exports the location of the system modules to make
variables, as well as the name of the default system module.
Test: TestClasspath in java_test.go, runs automatically as part of the build
Bug: 63986449
Change-Id: I27bd5d2010092422a27b69c91568e49010e02f40
linkageMutator removes srcs property of the shared variant of a lib in
order to reuse *.o files compiled for the static variant also to the
shared variant.
However, this causes problem when vendor-specific srcs are specified in
target: {vendor: {srcs: ["..."]}}. For example, let's assume
cc_library {
name: "libfoo",
srcs: ["foo.c"],
target: {
vendor: {
srcs: ["bar.c"],
},
},
}
Then,
static_vendor: inputs = foo.o, bar.o
shared_vendor: inputs = foo.o (from static_vendor), bar.o (from
static_vendor), bar.o
So, bar.o is included twice and this causes multiple symbol definition
error.
In order to handle the problem, vendor mutator is applied before the
linkage mutator and the vendor-specific srcs are squashed in the vendor
mutator.
Bug: 67731122
Test: build
Test: cc_test.go
Change-Id: I2a5390295dddfc41260e9b6f02746908cdf47228
If hostdex: true is specified for a java library, create an extra
Make module that copies the dex jar to a module with a -hostdex
suffix in the host output directory.
Bug: 67600882
Test: m -j checkbuild
Change-Id: I859dfaabeefdca714b566de94e00f74e03c85939
Putting a genrule dependency after a glob in a srcs property
was causing the generated files to be treated as globbed files
and hitting an interface conversion panic.
Bug: 67364649
Test: TestGeneratedSources in later patch
Change-Id: I8a076c9998fdd07e53769922be433a793ca575a9
This reverts commit 2370af0e23.
Reason for revert: New Build Breakage: aosp-master/aosp_arm64_ab-userdebug @ 4376965
Change-Id: Ibe4b819c4292457c454bf42e6d94fba3071ec04b
Once this is submitted and we don't have any more references to
target.linux, it can be added back as a common target for all
linux-kernel based targets.
Test: m nothing
Change-Id: Iae5e82eaed65f58950115f21530ae04afd0602b9
Arch specific properties were not being applied to modules
with OS set to Common.
Test: java_test.go
Change-Id: I8f1b49ca51b0cf96f78006dfcd121672e581d9c5
Add a properties to allow including files as resources, including
support for filegroups. Also add a flag that causes module sources
to be included in the final jar.
Test: java_test.go TestResources
Change-Id: Ida8ee59b28df9fe66952170f46470d3a09fd5d65
target.linux_glibc will apply to host builds with glibc, which is
identical to the current target.linux. In a future change, target.linux
will change to affect all targets using the Linux kernel (android,
linux_bionic, and linux_glibc).
target.bionic will apply to all OS variants using Bionic.
Bug: 31559095
Test: Add target.linux_glibc, target.bionic sections to an Android.bp, build
Test: m host
Change-Id: I677a67c22fba148fec264132311e355283f9d88d
Forcing sdk modules to be declared explicitly is unnecessary, just
add the required dependencies on the jar and aidl files.
Test: java_test.go
Change-Id: Ib28bdc1051c5825e7c0efb6adff1f9282675560e
Add TestArchConfig and NewTestArchContext to create a test context
that will run the arch mutator, which is necessary to test any code
that is inside a ctx.Device() block or similar.
Test: next patch
Change-Id: Ieb1a0326bc27fc18ba88b8d37655a7e0c6870d6c
Instead of trying to squeeze soong jars through prebuilt_internal.mk,
make a separate soong_java_prebuilt.mk.
Test: m -j checkbuild
Change-Id: Idadef3c21a6f618d573110f04c80d7db3db92031
The next change will cause java tests to fail if RUN_ERROR_PRONE=true
is set. Make tests more hermetic by faking an empty environment.
Test: soong tests
Change-Id: I01057cb0c42ec08ebed8aea87505dd01e9981915
Add a notice property to all modules which, if set, propagates to
LOCAL_NOTICE_FILE in make.
Test: m -j checkbuild, examine out/soong/Android*.mk
Change-Id: I565a5624dfd7b376b976b1a43dac5cea96869026
Pass the sdk_version property, the platform sdk version, or 10000
to dx as --min-sdk-version.
Test: m -j checkbuild
Change-Id: I5fae03f44153dc2d6244c33f4c055e746980aefe
Split InstallFileName into InstallExecutable that does chmod +x
after copying the file. Also remove InstallFile and rename
InstallFileName to InstallFile.
Test: m -j checkbuild
Change-Id: Id41ad4eafe521f6cd5d8cc250b7747ecb3da8dfc
All the stats below are for a hikey960-eng build on AOSP master.
Number of order-only inputs in out/soong/build.ninja:
Before: 2847162
After: 606508
Size of: Before After
out/soong/build.ninja 572MB 233MB -59%
out/soong/Android-hikey960.mk 15MB 9MB -40%
out/build-hikey960.ninja 367MB 358MB -2%
Ninja time in `m nothing`:
Before: 6.9s
After: 5.0s
Soong generation time:
Before: 11.6s
After: 6.6s
Test: m nothing
Test: wrote script that counted inputs, only duplicates were removed
Test: treehugger
Change-Id: I6b7c5ef5b1395014b7bf5fd0b8112d42bee127bf
O is released, so we don't get this knowledge from the development
codenames list any more.
Test: make ndk # with an API using introduced=O
Bug: None
Change-Id: I21104c5d97c89f2f85157ac7d73de8cd5a478571
Gingerbread is no longer supported. ICS is now the minimum target.
Test: make checkbuild
Bug: None
Change-Id: I896b83d9984d90d907fe518cc722a4a3722626a1
Right now this is hard coded in two places (here and in
core/binary.mk). Keep it in one place so it's easier to change.
Test: make checkbuild
Bug: None
Change-Id: I2a2c784d4c667b326f871e6144db92753c16c85f
Pass accumulated AndroidMkData to AndroidMkData.Custom handlers
and expose WriteAndroidMkData so that Custom handlers can write
out the normal make variables and then add their own.
Test: No change to out/soong/Android-aosp_sailfish.mk
Change-Id: Id9717132bbd6c5cf3af8596f3eaa9bbb05d98e40
It's never anything except nil, and it unnecessarily complicates
the implementations.
Test: m -j checkbuild
Change-Id: I3e3b7251f32ffa84dbdfd0448faf248c306ca808
It's never anything except nil, and it unnecessarily complicates
the implementations.
Test: m -j checkbuild
Change-Id: I021c7971ede3e11bbe08cf1601f4690ed4d1a036
It's never anything except nil, and it unnecessarily complicates
the implementations.
Test: m -j checkbuild
Change-Id: I8a117a86aa39aeb07d9d8d0686ef869c52784f19
Make the javac arguments match what is used by make, and export them
back to make. A future change will switch make to use the the
exported ones.
This makes a dx.jar compiled with soong have identical class files
as one compiled with make.
Test: manual
Change-Id: Ia5196f1f42bc564e99de22e32e72fd2930e9fbae
And make it work like bazel's java_import, using a "jars" property
instead of "srcs", and allowing multiple jars to be listed.
Test: soong tests
Change-Id: Ida2ace6412bd77b4feb423646000a1401004e0ea
Don't preserve the confusing names from make. Rename AllCodenames to
ActiveCodenames, add FutureCodenames, and add CombinedCodenames to
fetch the *real* AllCodenames.
To allow `introduced=P` in NDK library definitions, we need to know
that P exists. Use the combined list of current and future API
codenames generating stubs.
Test: make ndk
Test: check out/soong/api_levels.json
Bug: None
Change-Id: I435f9ce7446236edc268a84e33474044a55a6302
The java prebuilt tests weren't registering the prebuilts mutators,
which hid an issue where prebuilts modules weren't getting renamed
to avoid collisions with source modules of the same name, which
ended up causing the prebuilt module to try to add a dependency
on itself.
Test: java_test.go
Change-Id: I3327d27533591cb08a4b8b8b5e1668a6f72be8df
Without these, we'd need to go rewrite all the stub templates to stop
using the codename and use the number instead whenever an API is
released.
Test: changed something in libc to use introduced=L, make ndk
Bug: None
Change-Id: I1b8703655664c567f8a5c36219e73f11ba27d99a
Instead of calling SetNinjaBuildDir, pass it to bootstrap.bash, so that
the bootstrap package can set it consistently during bootstrapping and
normal execution.
Bug: 63720725
Test: m -j nothing
Test: mkdir o; ../bootstrap.bash; ./soong
Change-Id: Ica88d2d5f1461b5be49bfe6316c6ec4ef4d89d49
InitDefaultsModule and InitDefaultableModule were requiring the
module to be passed in as an argument twice, once as a
android.Module and once as an android.DefaultableMOdule. Use
a type assertion instead.
Test: builds
Change-Id: I57d721e68a4f016e02c17a678af5f2ba25a9a430
If owner is specified in a defaults module, we shouldn't concat that
with the owner specified in a module using that defaults module. A
string pointer will produce the correct behavior (overriding the default)
Bug: 37330627
Test: out/soong/Android-*.mk looks correct after this change
Change-Id: I64574e2ba81c11b042248d7a44702ec4534ee932
Add support for excluding paths from having integer_overflow applied to
them when using SANITIZE_TARGET=integer_overflow via an
INTEGER_OVERFLOW_EXCLUDE_PATHS make variable. This covers the soong side
of the change.
Bug: 30969751
Test: Build with SANITIZE_TARGET=integer_overflow
SANITIZE_TARGET_DIAG=integer_overflow
INTEGER_OVERFLOW_EXCLUDE_PATHS=<path> and confirmed this was no
longer being applied to binaries in that path.
Change-Id: I298b772f5425da28dff1cf007825be19558db3a8
Fix mutator registration for tests to allow different tests
in the same package to register different mutators.
Allow tests to track the resulting ModuleBuildParams objects
to use in assertions, and provide helpers for getting them.
For example:
config := android.TestConfig(buildDir)
ctx := android.NewTestContext()
ctx.RegisterModuleType(...)
ctx.MockFileSystem(...)
ctx.ParseBlueprintsFile("Android.bp")
ctx.PrepareBuildActions(config)
ctx.Register()
// Get the Inputs value passed to the javac rule for the foo module
inputs := ctx.ModuleForTests("foo".Rule("javac").Inputs
Test: java_test.go
Change-Id: I10c82967f5f3586d2c176f169906b571ed82fc73
device_kernel_headers is a built-in heder-only lib that provides device-specific
kernel headers. The header path is configured via a new product variable
SystemIncludeDirs, which is currently the mirror of
TARGET_PROJECT_SYSTEM_INCLUDES in the make world.
Note: generic kernel headers (bionic/libc/kernel) have been added to the
include path by default. "device_kernel_headers" module is for device-specific
kernel headers such as /device/*/*/kernel-headers.
Note 2: this is opt-in for Android.bp modules (i.e. header_libs :
["device_kernel_headers"] required.) while it is always provided to
Android.mk modules.
Bug: 62939405
Test: choosecombo to aosp_sailfish (or any other Pixel/Nexus targets)
BOARD_VNDK_VERSION=current m -j gralloc.msm8996 (or any other vendor
libs using vendor-specific kernel headers)
Change-Id: I81c60abc13942c89fff723d1544b27a81b300db0
Adds the SANITIZE_TARGET=integer_overflow build option to apply signed and
unsigned integer overflow sanitization globally. This implements the
Soong side of the build option.
An additional build option is provided to control whether or not to run
in diagnostics mode, controlled by SANITIZE_TARGET_DIAG. This works the
same way that SANITIZE_TARGET does and currently only supports
'integer_overflow' as an option.
A default sanitizer blacklist is added to avoid applying sanitization
to functions that are likely to exhibit benign overflows.
Bug: 30969751
Test: Building with and without the new flags, device boot-up, tested
various permutations of controlling the new flags from build files.
Change-Id: Ibc8a8615d3132f1a23faaf1cb4861f24c5ef734a
These should be install in /data/nativetest* with the rest of the tests,
but had been moved to /vendor/nativetest* accidentally. Add some tests
so that this doesn't happen again.
Bug: 63393698
Test: m -j blueprint_tools
Test: compare out/soong/Android-aosp_arm64.mk
Test: compare out/soong/build.ninja
Change-Id: Id3b08a7e3908955df18a043a02ea576dc88086c3
cc_defaults {
name: boo,
device_supported: false,
}
cc_library_static {
name: foo,
defaults: [boo],
}
Soong still tried to build foo_android_arm/arm64 device target which is not what we
expected.
Test: m -j checkbuild
Change-Id: I26a67c9ea024f5458f0818def0fa10cecc5fb7cf
Change module factories from returning a blueprint.Module and a list
of property structs to returning an android.Module, which holds the
list of property structs.
Test: build.ninja identical except for Factory: comment lines
Change-Id: Ica1d823f009db812c518f271a386fbff39c9766f
Running prebuilts/build-tools/build-prebuilts.sh in a repo that
contains art/build/art.go panics because config.Targets[Device] is
empty. Check the length before accessing the slice.
Test: prebuilts/build-tools/build-prebuilts.sh
Change-Id: Ifb9fe0fad07b22d6b574f505c08c5c761278aad0
This causes Soong to put the outputs of each genrule into a temporary
location and copy the declared outputs back to the output directory.
This gets the process closer to having an actual sandbox.
Bug: 35562758
Test: make
Change-Id: I8048fbf1a3899a86fb99d71b60669b6633b07b3e
This CL fixes a bug with how Soong builds the install path for native
tests. Tests are installed in /data/nativetest* by default, but the
logic was wrongly redirecting sanitized versions to
/data/asan/system/nativetest*. With this fix, they are correctly
redirected to /data/asan/data/nativetest* instead.
Bug: 37942061
Test: lunch marlin_asan-userdebug && \
SANITIZE_TARGET="address" m -j70 \
# nativetests are generated in /data/asan/data/nativetest*
Change-Id: I0c132af5d443151c44219d231770049fddf79bbe
Some of ART tests fail to mmap memory unless
LibartImgDeviceBaseAddress() is further lowered.
Test: booted MIPS32R2 in QEMU
Test: booted MIPS64 (with 2nd arch MIPS32R6) in QEMU
Test: test-art-target-gtest
Test: testrunner.py --target --optimizing
Test: repeat all of the above in configurations
ART_READ_BARRIER_TYPE=TABLELOOKUP,
ART_USE_READ_BARRIER=false
Test: repeat the above tests on CI20
Change-Id: I48cd9d8074e07073cc6abb2153660e00017b9e6b
We now add export_static_lib_headers, export_generated_headers to the
filters while dumping the abi of a library using header-abi-dumper
(through -I<dir> additions to the invocation of header-abi-dumper and
header-abi-linker)
Also add support for zipped reference source based abi dumps.
Test: mm -j64 in hardware/interfaces/nfc/default/1.0 produces
android.hardware.nfc@1.0.so.lsdump with abi filtered out using generated
headers.
Test: Copied the linked abi dumps produced by mm -j64 in bionic/libc to
prebuilts/abi-dumps/ndk and gzipped them. Ran mm -j64 again in
bionic/libc and verified header-abi-diff getting invoked.
Bug: 32750600
Change-Id: I26210af908c87a6143e39fa25f50307acb68a387
The pdk product variable will be set by make when TARGET_BUILD_PDK
is set, and can be used to disable modules that won't build in the
PDK.
Test: builds
Bug: 62086238
Change-Id: I2191a57b9b06680fd741308e7464275a89773530
This exposes PRODUCT_FULL_TREBLE.
This is required so that we can add sane defaults to Treble code.
Test: manual
Bug: 38415912
Change-Id: Ia3ea4aeee799b998380a1600026075dab66286a3
Added params to WriteFile rule that allow users to turn on the
specific 'echo' options, such as "enable the interpretation of backslash
escapes".
Bug: b/38261000
Test: tested with real data files.
Change-Id: I5145056b5c2522a3a72b9436055ec95d2fbd94c0
This reverts commit 4068a5db6c.
Now the Mac xcode-select and xcrun goes through Config.HostSystemTool,
which will grab $PATH through Config.Getenv
Test: m -j (on mac)
Change-Id: I2632c4fdb2ec961e59944cf02ff165e0fd3c869d
Descriptions currently look like:
[ 0% 4/29328] cc out-soong/.intermediates/external/clang/lib/Sema/libclangSema/android_arm_armv7-a-neon_denver_static_core/obj/external/clang/lib/Sema/SemaCodeComplete.o
This is not very helpful - most of the characters are used to show the
output path, which contains useful information like target architecture,
but also contains most of the path to the source files twice, and less
useful information like the exact variant name used by soong.
Make the descriptions look like:
[ 0% 3/29329] //external/clang/lib/Sema:libclangSema clang++ SemaTemplateInstantiate.cpp
This is //path/to/module:modulename tool relative/path/to/source/file
Test: builds, looks pretty
Change-Id: I3087aa7d4eb1860ef6239d77407b8b35445616d7
This reverts commit e78473c33c.
Reason for revert: Broke mac builds, which rely on $PATH to find xcrun
Change-Id: I58711b94d472ab828823afaf53394a1aa869f3f5
Store the os Environment on init, then clear it so that it's only
available via a call to Config.Getenv(). That way we're guaranteed to
know about all dependencies.
Test: Add some os.Getenv/Config.Getenv calls and ensure correct behavior
Test: cs/ shows no os.Getenv / os.LookupEnv / os.ExpandEnv calls that
would be affected
Change-Id: Ic519e0c871f340e752df03b8af0599d47c1d91d8
Not plumbed through yet, but will at least prevent us from losing
track of which modules will not build in the PDK.
Test: builds
Change-Id: I4c1a0e39104b104dc1ff23c4391d5e83b3002d3a
Support using strings as product variable substitutions, and
add tests for printfIntoProperty.
Test: varaible_test.go
Change-Id: I06cfadfb1d3fc81da72fb71323706df20426c8b7
The base module handles all the common functionalites, such as version
compatibilty check, version variations split, source file format check,
source/data file duplicate check.
The library/binary module focuses on how to generate binary build actions,
such as setting up stub script, zipping, filling in __init__.py in
runfiles dir tree.
Bug: b/31676493
Test: go test under python package
Change-Id: I06608369f350f7195873d459e1c8d1bdb811e77e
TARGET_USES_HWC2 is going away, but propagate it to soong for now
to unblock converting dependency chains that end in libhwui.
Test: soong tests
Change-Id: I20f1269caea1b5f5fc39239daa6b2e610bd36eb9
Turn on installation in the mega device build, it is necessary for the ndk
sysroot installation. Partially fixes mega device builds, they also need
to run without -w dupbuild=err on the ninja command line because multiple
variants of the same architecture try to install to the same ndk sysroot.
Test: mega device build
Change-Id: I982d77f9ff19f5bc29fc9fe54a0df8db3579c3e3
Current modules must use armv7_a_neon to specify source files that
compile only with neon. If a future arch variant also supports neon,
all these modules will fall back to non-neon. Support a neon arch
feature that modules can use instead. Similarly, support dspr2 for
mips.
arm_device.go was also mixing armv7-a-neon with armv7_a_neon. Use
armv7-a-neon consistently, and fix the - to _ when creating the
property structs.
Test: m -j checkbuild
Change-Id: I24d3764280ab3bcbb9a73c0934edc9b99fc7f6a0
header-abi-dumper: dumps abi exported by source files for Vndk.
header-abi-linker: links abi dumps produced by header-abi-dumper.
header-abi-diff: compares linked dumps.
Test: mm -j64 showcommands > make_log in bionic/libc.
This produced linked dumps in out/soong/.intermediates.
Copied these dumps to
prebuilts/abi-dumps/ndk/current/arm64/source-based/.
Changed the abi and re-ran mm -j64 showcommands > make_log
confirmed that the build reported compatibility breakge without
actually failing (advisory mode).
Change-Id: Iccad6908fe68a80f47230751671d156893b96ead
Bug: 37110010
Test: See related change - verified manually that the default init_rc
file as well as the 'debuggable' init_rc file were executed.
Change-Id: Id6a51201cf10633e9a86d2c76a3e64b50225bf55
This CL changes the ENABLE_CFI flag to default to enabled. Setting it
to false will override local settings to enable CFI.
Bug: 30227045
Bug: 22033465
Test: m -j40 works and device boots
Test: cfi is honored unless the global flag is set.
Change-Id: Ie3285c5eac60c8f6012c6b6c23be149a8787af0c
Enable arm_on_x86 mode whenever compiling for x86 on the device,
and either arm is listed as an ABI on the x86 arch, or arm exists
as a target arch.
Bug: 35286489
Test: examine bcc cflags
Change-Id: Iebd0e7b95f584d25773a60474c27425cac7a578e
The vendor image will have more than just proprietary modules in it
under Treble, so let's stop marking open source code as proprietary just
to move it to vendor.
Bug: 36452052
Bug: 37134596
Test: compare build.ninja before/after, no changes.
Test: Set vendor: true, ensure it works.
Change-Id: I44b0ec7007d0e311bdcbd44b238b1ef2d05cc6ff
(cherry picked from commit aa118f95a6)
The vendor image will have more than just proprietary modules in it
under Treble, so let's stop marking open source code as proprietary just
to move it to vendor.
Bug: 36452052
Bug: 37134596
Test: compare build.ninja before/after, no changes.
Test: Set vendor: true, ensure it works.
Change-Id: I44b0ec7007d0e311bdcbd44b238b1ef2d05cc6ff
Instead of having a module define `use_vndk: true`, assume that we're
building with the VNDK if we're installed on the vendor partition, and
BOARD_VNDK_VERSION==current. This now matches our behavior in Make.
Once BOARD_VNDK_VERSION!=current, we'll need to disable modules that
need to otherwise compile against the VNDK, since we can only compile
against the current VNDK.
Test: build.ninja for aosp_arm is the same before/after
Test: Ensure there are no boards that set BOARD_VNDK_VERSION
Change-Id: If937fa7bdb119648137af52daebadf486163484b
(cherry picked from commit 11b261472a)
* Use mcpu=cortex-a57, since it is the closest to Kryo.
* Clang doesn't support Kryo as a target yet.
Bug: 36728278
Test: Built and booted a sailfish using Kryo.
Change-Id: Ic0c9588d86fba41896e50e3f0cf0d2b310ffee93
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
* Kryo is closer to Krait than to cortex-a53, so choose Krait defaults.
* Made together with Jake Weinstein (xboxfanj)
Bug: 36728278
Test: Built and boot sailfish set up as kryo.
Change-Id: I04fa0a1dca6f97ae19202d28ee3ce8a59bf169b5
Signed-off-by: Alex Naidis <alex.naidis@linux.com>
This CL moves the location of ASAN-ified libraries on disk in the
following manner:
/data/lib* --> /data/asan/system/lib*
/data/vendor/* --> /data/asan/vendor/*
There are a couple of advantages to this, including better isolation
from other components, and more transparent linker renaming and
SELinux policies.
(cherry picked from commit 1dd8839286)
Bug: 36574794
Bug: 36674745
Test: m -j40 && SANITIZE_TARGET="address" m -j40 and the device
boots. All sanitized libraries are correctly located in /data/asan/*.
Change-Id: I06bf459260ec451d4495a02562f640ad622f34c2
The NDK stub library generator needs to know what unreleased API
levels exist and what the ordering is between them. This singleton
will later be expanded to generate api-level.h to cut down on
duplication.
Test: make out/soong/api_levels.json && cat out/soong/api_levels.json
Bug: None
Change-Id: I53126dd2cacb67c331c44f7d7c77c98b176b93cd
This CL moves the location of ASAN-ified libraries on disk in the
following manner:
/data/lib* --> /data/asan/system/lib*
/data/vendor/* --> /data/asan/vendor/*
There are a couple of advantages to this, including better isolation
from other components, and more transparent linker renaming and
SELinux policies.
Bug: 36574794
Bug: 36674745
Test: m -j40 && SANITIZE_TARGET="address" m -j40 and the device
boots. All sanitized libraries are correctly located in /data/asan/*.
Change-Id: I06bf459260ec451d4495a02562f640ad622f34c2
Phony module was broken since it reached the code line to try to
invoke some specific arch functions which caused the runtime panic
error.
Bug: 36739664
Test: mm -j32 -k
phony {
name: "shell_and_utilities",
required: [
"grep",
"gzip",
"mkshrc",
"reboot",
"sh",
"toolbox",
"toybox",
],
}
in system/core/shell_and_utilities/Android.bp
Change-Id: Idf7da17d431aae5f0c56b08f5e5eef14ae4eef50
Instead of having a module define `use_vndk: true`, assume that we're
building with the VNDK if we're installed on the vendor partition, and
BOARD_VNDK_VERSION==current. This now matches our behavior in Make.
Once BOARD_VNDK_VERSION!=current, we'll need to disable modules that
need to otherwise compile against the VNDK, since we can only compile
against the current VNDK.
Test: build.ninja for aosp_arm is the same before/after
Test: Ensure there are no boards that set BOARD_VNDK_VERSION
Change-Id: If937fa7bdb119648137af52daebadf486163484b
The concurrent copying GC allocates large contiguous
space at low addresses. This moves the image(s) out
of the way to higher addresses.
Test: booted MIPS32R2 in QEMU
Test: booted MIPS64 (with 2nd arch MIPS32R2) in QEMU
Change-Id: I1c54e6142844ab1ad7820765ae313650e9ab8875
* changes:
Add support for binary and static library and prebuilts
Rename to cc_prebuilts_library_shared
Fix installing prebuilts with no source module
Support owner properties in Android.bp files and propagate the value
to make using LOCAL_MODULE_OWNER.
Bug: 36413696
Test: builds
Change-Id: Iac1cf7f2a14c9e3e6d85d6e09d5adb08a7078d00
Prebuilts that were not overriding a source module would never
get UsePrebuilt set, which would cause SkipInstall() to be called,
and prevent the module from being exported to make.
Test: prebuilt_test.go
Change-Id: Ib09300efff26f09245331e9919ae2db82b0c3625
This reverts commit 4c48f724e1. It's
causing test failures, warnings and complaints, so backing it out and
we'll resolve those before putting it back in.
Bug: 33681361
Test: m -j
Calling android.NewContext() in tests results in a context that
contains all the mutators, module types, and singletons, which
causes unexpected interactions in unit tests. Create an empty
context instead, and add in only the necessary mutators, module
types, and singletons.
Bug: 36366816
Test: soong tests
Change-Id: Ic61262c37e3436b3ad4ccaca18b737021c304be6
This adds a toolchain definition for LinuxBionic that only supports
Clang/64-bit. It pulls pieces from the x86_linux_host and x86_64_device
configs, and uses the android clang triple, with some manual overrides.
To enable building this, set your soong.config file to:
{"Host_bionic": true}
Bug: 31559095
Test: out/soong/{Android,make-vars}-aosp_arm64.mk the same with or
without host bionic turned on
Test: No change to out/soong/build.ninja before/after this change
Change-Id: Id97dda8bd9aa670c32aed31fbe6aaa8175e70b59
This breaks when I enable a second host configuration for host bionic.
Make non-arch modules !Device and !Host by adding "Generic" as the zero
value. Make the SkipDeviceInstall config only apply to Device modules.
Test: out/soong/build.ninja is identical expect for comments
Change-Id: I6e2bd3814a9968554d3901a508d822d5c79d08a0
Shared libraries are now installed to different directories depending on
their types.
* NDK libraries: /system/lib/ndk
* VNDK libraries: /system/lib/vndk
* VNDK-ext libraries: /system/lib/vndk-ext
* Framework-only libraries: /system/lib
* Vendor-only libraries: /vendor/lib
* Same-process HALs: /vendor/lib/sameprocess
In addition, a new module type vndk_ext_library is added. It is almost
identical to cc_shared_library but it introduces another attribute
'extends'. This is use to reference the vndk library that this vndk-ext
library is extending.
For example, in order to extend a vndk library libFoo:
cc_library {
name: "libFoo",
srcs: [...]
}
---------------------
vndk_ext_library {
name: "libFoo-extended",
srcs: [...]
extends: "libFoo"
}
Then, libFoo will be installed as /system/lib/vndk/libFoo.so and
libFoo-extended will be installed as /system/lib/vndk-ext/libFoo.so.
Note that file name of the latter is libFoo.so, not libFoo-extended.so:
file name of an extending module is automatically set to that of the
extended module.
Bug: 33681361
Test: build & run. Libraries must be in the correct directories.
Change-Id: Ia1eb3940605d582a252c78da0f3a5b36fdab062b
Directly using blueprint.BaseDependencyTag might cause issues because
if other modules also use blueprint.BaseDependencyTag, the
different type of modules will be mixed when fetched based on Tag.
Test: m -j out/soong/build.ninja
Change-Id: I290ca2eea063e5daa4f572985b105b958fa1d0fd
Specify list of paths to exclude from coverage instrumentation.
Test: make NATIVE_COVERAGE=true COVERAGE_PATHS=hardware/interfaces
COVERAGE_EXCLUDE_PATHS=hardware/interfaces/graphics
Bug: 35769817
Change-Id: I3bf10e5e5697d140d6cff73d000768b00aa28ca4
This is configured the same as make -- a global NATIVE_COVERAGE=true
flag to allow native coverage, then COVERAGE_PATHS=path1,path2,... to
turn it on for certain paths.
There are .gcnodir files exported to Make and saved in $OUT/coverage/...
files which are `ar` archives containing all of the compiler-produced
.gcno files for a particular executable / shared library.
Unlike the Make implementation, this only passes links the helper
library (automatically through --coverage) when one of the object files
or static libraries being used actually has coverage enabled.
Host support is currently disabled, since we set -nodefaultlibs, which
prevents libclang_rt.profile-*.a from being picked up automatically.
Bug: 32749731
Test: NATIVE_COVERAGE=true COVERAGE_PATHS=system/core/libcutils m -j libbacktrace libutils tombstoned
$OUT/coverage/system/lib*/libcutils.gcnodir looks correct (self)
$OUT/coverage/system/lib*/libbacktrace.gcnodir looks correct (static)
$OUT/coverage/system/lib*/libutils.gcnodir doesn't exist (shared)
$OUT/coverage/system/bin/tombstoned.gcnodir looks correct (executable)
Test: NATIVE_COVERAGE=true COVERAGE_PATHS=external/libcxxabi m -j libc++
Confirm that $OUT/coverage/system/lib*/libc++.gcnodir looks correct (whole_static_libs)
Change-Id: I48aaa0ba8d76e50e9c2d1151421c0c6dc8ed79a9
The "required" field within "phony" module in .bp file has to be
non-empty. And "phony" module doesn't contain any "src" files.
Bug: b/32981153
Test: make -j out/soong/build.ninja
Change-Id: Iaa2009467c44702572b97ca9cbd02c1d368c8d83
If Config.GetEnv was called when envDeps was uninitialized (for
example in a test) it would panic, which if recovered (for example in
a test) would cause it to continue without unlocking the mutex, and
could later deadlock. Fix the initialization by initializing in
GetEnv if necessary, and use defer to avoid holding the mutex after
a panic.
Test: soong tests
Change-Id: I453522faaf47ff6fbc4702345cfe177100bdc522
Allow tests to specify a data property that lists files or filegroup
modules that will be packaged alongside the test. Also add a path
property to filegroup modules to allow shifting the path of the
packaged files, and add ExpandSourcesSubDir to expand the filegroup
sources while including a shifted relative path in the Paths objects.
Test: soong tests, manually adding data to a module
Change-Id: I52a48942660e12755d313ef13279313361b4fc35
ModuleSrcPath contains an embedded basePath as well as a
SoucePath that contains another basePath. Remove the embedded
basePath, and make the SourcePath embedded.
Test: no change to build.ninja
Change-Id: I3cdf3477eca41ed35fac08a892aab22cbcdb2224
This CL ensures that locally enabling CFI in a .bp file is not honored
unless it is enabled globally using ENABLE_CFI=true first, effectively
hiding it behind a flag.
Bug: 30227045
Bug: 22033465
Test: m -j40 works and device boots
Test: cfi is correctly honored only when the global flag is set.
Change-Id: Iccc6e4bf5e7828ab8ce6056f3e40922712faa0d8
build/make/board/generic_mips/BoardConfig.mk sets
TARGET_CPU_VARIANT := TARGET_ARCH_VARIANT, which can cause the same
variant-specific properties to be applied twice.
Test: lunch aosp_mips-eng && mmma -j external/skia TARGET_ARCH_VARIANT=mips32r2dspr2-fp
Change-Id: Ie4abe804f8ce898bf89664ae749b9bc8ac3e649f
filegroup is a module that contains a list of files, and can be used
to export files across package boundaries. filegroups (and genrules)
can be referenced from srcs properties of other modules using the
syntax ":module".
Test: m -j
Change-Id: I3d6fc4819c0b4225b474e0ad42f0d947f55a5961
Allow Android.bp files to set properties when an x86 device supports
arm abis.
Test: lunch aosp_fugu-userdebug && mmma -j frameworks/compile/libbcc
Change-Id: Ic142a3f93175924a826dea9908b2e5183a486184
The eng product variable will be true for builds where
TARGET_BUILD_VARIANT==eng.
Test: mmma -j frameworks/compile/libbcc
Change-Id: I808129740643495370bea64df3fe5bcf01eb9915
In lieu of having an environment variable to override prebuilts
preferences (for now), fall back to prebuilts whenever the source
module is disabled. Complicated source modules (like LLVM) that
already have Go logic can then use an environment variariable to
disable their source module whenever they want to use prebuilts.
Computes UsePrebuilt once using information from both the source
and prebuilt modules, and stores the result in the prebuilt module.
Test: m -j libLLVM
Test: m -j libLLVM FORCE_BUILD_LLVM_COMPONENTS=true
Change-Id: Ib0819a03e9492e55f55de3402e05b3a043644a39
If the source module is being used instead of the prebuilt module,
disable installing the prebuilt module.
Test: m -j checkbuild
Change-Id: I55e77021b0f9572b0737d960cba89274f696775d
Move the calls to the AndroidMkProvider to after checking if the
module is disabled. Disabled modules may not have all the state
necessary to generate the data.
Test: m -j libLLVM
Change-Id: Iab27a26e604f3810bcc2d9f61b1839aa7bc428b4
Export a list of libraries in the VNDK, which is currently empty.
Take in Make's global BOARD_VNDK_VERSION, and use that as the SDK
version for modules that specify use_vndk: true.
Modules that use the vndk have some configuration as if they were
building against the NDK (the absence of globally defined headers), but
in other cases look like platform modules (using the platform libc++,
for now).
This change does not attempt to enforce any linking constraints, that
will come in a later patch.
Test: out/soong/build.ninja doesn't change
Change-Id: I3be206b67015ac5177b7eef4a451c579e3dc903f
If a genrule sets deps_file: true, use a file in the gen directory
as a GCC-style deps file.
Test: m -j libLLVMObject
Change-Id: Id410165847e4eaea1853a392512e38787c431523
An escaped $ sign in an expanded string needs to be re-escaped so it
can pass through to ninja.
Test: m -j libLLVMObject
Change-Id: I2f7890417fd6f534e6e21cace3d86ae9ef708ce5
Currently our only default-disabled Os is Windows since it's HostCross,
but we'll be adding non-default Host and Device types in the future.
Bug: 31559095
Test: out/soong/build.ninja is identical
Change-Id: I2bc3a3cc76f2c95ea040bc34ba6706fcc178c68d
N has merged into aosp now.
Test: OUT_DIR=ndk-out DIST_DIR=ndk-dist \
build/soong/scripts/build-ndk-prebuilts.sh
Bug: None
Change-Id: I2685c700e2cb138708a3e8a5f51281c30fb582db
To allow genrules with more than one tool, rename the tool property to
tools and make it an array, replace $tool with $(location <label>),
and use $() for other variables for consistency.
Also remove the host bin directory from the genrule path, and the
$srcDir variable, using either of them would have caused dependency
issues.
Bug: 31948427
Test: compare build.ninja
Change-Id: Icf6d3bce2bea00fec1363fd65c0bdf96d09281bf
(cherry picked from commit de6bd86d24)
Move Soong's globbing-with-dependencies support into Blueprint so it can
be used for subdirs= lines in Android.bp files.
Blueprint has a slight change in behavior around subname= lines, it now
always uses the subname and doesn't fall back to Blueprints. To support
the Blueprints files in build/blueprint, use them directly with build=.
Test: build, add source file that matches glob, rebuild
Change-Id: Ifd0b0d3bc061aae0a16d6c7ca9a1cd8672656b4d
soong_build is run twice simultaneously now, once for manifest
generation and once for docs generation. If one starts writing the
default soong.config file, and the other starts reading it, the reader
can see an empty file and fail. Write the soong.config file to a
temporary file and the atomically rename it into place.
Bug: 32628314
Test: rm out/soong/soong.config && m -j blah && cat out/soong/soong.config
Change-Id: I8119b11d45093284b24cbc926d81eb9ea4bf2e27
We were emulating this for proto files, standardize it and make the
other generators use it as well.
Test: Compare out/soong/build.ninja before/after change
Test: mmma -j system/tools/hidl
Change-Id: I1888c7b981749060a398387bbb9b481270bf6d75
Allow specifying one or more architectures to sanitize for. Defaults to
sanitizing all architectures. Allows reducing the cost of sanitization
by only sanitizing half of the platform.
Bug: 29498013
Test: No change to build.ninja files with m -j SANITIZE_TARGET=address
Test: m -j SANITIZE_TARGET=address SANITIZE_TARGET_ARCH=arm
Test: nm -D $OUT/system/bin/app_process64 | grep asan, no asan symbols found
Change-Id: I972cccc2f19443895d47b44bd8104105f93ffb7d
For every file which we can run clang-tidy (C/C++ clang-built), we add a
new build node that depends on the object file (since clang-tidy does
not export a depfile), and is depended on by the link step. This is
better than how we're doing it in make, since calling tidy can be turned
on or off without needing to rebuild the object files.
This does not attempt to port WITH_TIDY_ONLY from Make, since the way
that it works is broken (due to the lack of a depfile).
Bug: 32244182
Test: WITH_TIDY=true mmma -j bionic/libc
Test: ./soong (Setting ClangTidy: true)
Change-Id: I40bbb5bb00d292d72bf1c293b93080b5f9f6d8ea
This field points to the license file for the headers being shipped.
Test: make ndk && less $SOONG_OUT/ndk/NOTICE
Bug: None
Change-Id: I386f4e6f6d9776f422ffc09b8dab69e1911b08a4
Use a temporary directory as the build directory during tests so files
don't get written to the source tree.
Also add a few more tests for prebuilts with no file specified.
Bug: 31800129
Test: m -j, make sure .soong.environment is not written to the source tree
Change-Id: I623bc114b2ff534c8df9fb3ce273e804711f8f05
Support prebuilt shared libraries as an initial proof-of-concept of
prebuilts. Future changes will support binaries and static libraries,
and the ability to select which to use based on something besides
blueprint properties.
Test: TestPrebuilts run during m -j
Change-Id: I6e84da667e9005ae11844bad01d25cbe4ced1ce3
Mutators are registered into a global variable, and then into the
context when it is created. Only call registerMutators once so that the
same mutator is not listed multiple times in the global variable when
multiple contexts are created.
Test: go test -v android/soong/android
Change-Id: Ie9e3ed09a89b848462b898476cdfb81a90c64bd3
Blueprint is going to abdicate responsibility for determining the name
of a module. Add a name property, and a method to retreive the name.
Test: build.ninja identical
Change-Id: I09c6f5283cd6e28ad4b04c24c5ab8b00f71ae2ab
Register mutators inside lambdas that are called in a defined order to
correctly order mutators before and after the arch and deps mutators.
Test: build.ninja identical
Change-Id: Iefe2a3515aee8570e76a6e76925db4cda0e9e822
Mutator registration is tightly coupled with the android package, move
all registration from the soong package to the android package.
Test: build.ninja identical
Change-Id: Ie183d0b52cc7431c9e05b231934d189208ef1efe
Using new .KATI_READONLY extension.
Test: build/tooks/kati_all_products.sh on AOSP master and internal
Change-Id: I71794c1268456c3403a351c1b242f1e78b060355
The google-benchmark library is disabled on Darwin, so we cannot create
host benchmarks. Instead of having every user specify this, put the
logic in Soong. Then if we decide to support it later, it's an easier
switch.
Test: build.ninja identical before/after on Linux
Test: Ignores failing cc_benchmark_host on Darwin
Change-Id: I61f3a571fd160d8e479a512992bc68601f1c9b28
The toc optimization had been disabled for host builds to ensure that
the timestamp of the final binary changed whenever its implementation
changed, in order to support rerunning host tools that were modified
during incremental builds. However, only the final install rule must be
re-run to update the timestamp, and not the link rule.
Update the shared library install dependencies to use normal
dependencies instead of order-only dependencies for host modules, and
then enable the the toc optimization for host modules. If the
implementation of a library changes it will be reinstalled, and
libraries or binaries that depend on it will also be reinstalled.
Also move toc generation to happen on the packed, stripped library,
which is what will be used for linking, to ensure that it is available
at link time when depending only on the toc file.
Bug: 26015464
Test: m -j; touch system/tools/hidl/Annotation.cpp; m -j, verify
out/soong/host/linux-x86/bin/hidl-gen is updated
Change-Id: I8953261d2209376f3dccbf0b1a93f7af4e45c4d0
Allows changing cflags based on the build variant being `eng` or
`userdebug`. This is used by some modules to turn on asserts for these
builds, and used by adb to allow root on non-user builds.
Test: Use this flag in an Android.bp, check the ninja file.
Change-Id: I27a5081378e94920482b4a742d65c46065047573
The soong checkbuild target is created by the last varaint of each
module. Don't return early if the variant is disabled, which would
prevent creating the checkbuild target.
Test: m -j checkbuild
Change-Id: Ib260bae7fb4be7255f5c1b34dfbd50225f308005
Replace PropertyCustomizer with a more extensible hooks mechanism.
Instead of passing an object that satisifies an interface, pass a
function pointer that takes a context interface as an argument. Callers
can use lambdas to capture any other necessary parameters like property
structs.
Also add two new hooks, arch hooks that occur after splitting modules
into arch variants, and install hooks that occur each time a file is
installed to the output directory.
Change-Id: I3a3e34aa97f1a92d3a31e5004b4b1ba68869d242
Windows builds have the first arch set to x86 and the second set to
x86_64, and always prefers 32-bit. Specify the 32-bit preference
manually, instead of relying on the architecture order for now.
Change-Id: I1900ec095e7773f68c0db293bfc48bd0815661b2
Track the primary architecture selected for each class based on the
module's multilib setting and the global config. Fixes building
binaries with multlib set to first and DevicePrefer32BitExecutables set,
and fixes symlinks to binaries with multilib set to prefer32.
Bug: 31452121
Test: mmma -j art HOST_PREFER_32_BIT=true
Change-Id: I75094df42f3273f6d613e4058eaa565957174c28
Allow multilib to be set to "prefer32", which will compile as 32-bit if
available, otherwise as 64-bit.
Add target.device.compile_multilib and target.host.compile_multilib. If
set, they override the top-level compile_multlib property.
Change-Id: If658a035b5f441547bc74526feb1d34f773776ff
HOST_PREFER_32_BIT was used during the switch to 64-bit host tools to
keep the SDK building as 32-bit, but is never set any more.
Change-Id: I7c2db269d3f7fa1f4e0c022cbced789755a62d81
When the UseGoma flag is set, put all rules except the C compilation
rule in an externally defined local_pool, which will have been created
by kati. The gomacc wrapper will already be in the CC_WRAPPER
environment variable.
Bug: 31142427
Change-Id: I699d4edff2e302eee398dad8692ceb14721a628c
Add a symlink_preferred_arch property to binaries to allow compiling the
binary for multiple architectures and then creating a symlink to the
preferred archicture, for example dalvikvm32 and dalvikvm64, with
dalvikvm symlinked to dalvikvm64.
Test: mmma -j art/dalvikvm
Change-Id: Ied15f2be9d52c01006fe8ac207c175b78558eab1
Allow any module factory to insert a PropertyCustomizer on the module,
which will be called before any other mutators. The PropertyCustomizer
can append or prepend to any properties, allowing module types to extend
other module types by modifying the public, stable interface provided by
the properties.
Change-Id: Idff02be80d939a70df1c6bbccffdd1f04ff975d2
Add DeviceConfig to store per-device configuration information. Put a
OncePer object inside Config and DeviceConfig, which computes a value
once per key per object to allow build logic to store arbitrary
per-build or per-device computed values.
Change-Id: I1a38b426f29d223ef5e803e0d4d9604500de2fd2
This adds LOCAL_REQUIRED_MODULES to the generated Android.mk for any
dependencies in the "required" list of a module.
Change-Id: Ibcfe89dce7d247eb441af94e37388a59e71a75a9
Append .Parallel() to the defaults RegisterTopDownMutator call to tell
Blueprint it can run it in parallel. Saves ~500ms in soong_build.
Change-Id: I43ddd9d6995674ccc06fed6928514f15a15712c1
Append .Parallel() to all of the RegisterBottomUpMutator calls to tell
Blueprint it can run them in parallel.
Test: identical build.ninja, passes race detector
Change-Id: I969a0689522d4cba7c8ff51e2aa00fe2fd338a89
BottomUpMutators are going to lose their ctx.Visit* functions in order
to allow parallelizing them, move defaults to a TopDownMutator using
WalkDeps to only visit defaults modules.
Test: no changes to out/soong/build.ninja
Change-Id: I54ba65a7e2ae9503f4d217f63aa9178a7c5341f0
Blueprint can now handle creating a zero-valued struct when it
encounters a nil pointer to a struct, reducing the amount of allocations
and copying for arch structures that are not used.
Change-Id: If36da5603dbe6469fe6406c821f21a122ea6ea49
To allow cflag changes when using the "old" binder API that are still
used by all 32-bit only configurations. We can't just use
android32/android64 here, since some devices use a 64-bit kernel with a
32-bit userspace, which is configured as "android32".
Change-Id: I8450484b75d59c1855c7ba36260c08925e7b28ad
Make will combine this information with statistics gathered from the
Make build, then dump it out to a build artifact.
Change-Id: I2bd84d4f8a8d145c37e3a4c4fcc8fece99e2a1de
The NDK stub libraries need to use subname, but should use "." instead
of "_".
Test: Still builds.
Change-Id: Ib4b0303531e03968c55671ea167ab987adebd783
Soong delegates installation to make when embedded in make, so it can't
create any dependencies on installed files. Only create the symlink
rules if not embedded in make.
Change-Id: I26d2857e6d544b963a0c52145a8666ba30864838
Allow modules to specify a list of names to create alias symlinks, and
pass the list to make so it can create them. Also pass
LOCAL_FORCE_STATIC_EXECUTABLE for static binaries so that make can avoid
a linker_asan -> linker -> linker_asan dependency loop.
Change-Id: I314d088095ac5f43641ed2cf8247c14c27e23b93
These are used by libcutils, and are equivalent to the ENABLE_CPUSETS
and ENABLE_SCHEDBOOST make variables.
Change-Id: I65d0eed49b5bcb4d6415e9e64a0159209a43b922
When collapsing properties for applying defaults, bool is ORed, *bool is
replaced, which is the behavior we want here.
Change-Id: I40ac5035bedcd4b1bbf50e054f8527523f9f6f79
There are some builds where we only want to build host tools, and we may
not have a valid device configuration. Support 0 device architectures
and modules that don't have any configured way to build.
Change-Id: I732251e1cd3cdbfafbc6ef6a550e33e653876f7c
cc.ModuleContext.module() returns a *cc.Module, and is left over from
when the cc package tried to use inheritance. Remove it and the last
few users.
Change-Id: I9b42ca59689c1b0ada7980fbec923747ed3a53d3
libz has a case where the static and shared versions of the device
library are supported, but the host shared library must not be defined,
since it would conflict with the system provided version. There's a
separate module 'libz-host' that provides the host shared library
implementation.
So extend the 'static'/'shared' properties to include 'enabled', which
uses arch_variant so that it can be mutated per-arch. It cannot override
a top level 'enabled: false'.
Change-Id: Ia0f1ff31ad77dc4cb148c531b70158e3245031d5
sanitize.begin() modifies the slice returned by Config.SanitizeHost()
and Config.SanitizeDevice(), return a copy so the global slice doesn't
get corrupted.
Bug: 29188876
Change-Id: I4c7a59b7e96529166519b23a1ebda39b8e060c58
Soong's multi-architecture building has grown complex, with the
combination of HostOrDevice+HostType+Arch necessary to determine how to
build a variant of a module, and three separate mutators to split each
into its variations.
Combine HostOrDevice+HostType into Os, which will be Linux, Darwin,
Windows, or Android. Store Os+Arch as a single Target.
Change-Id: I92f2e2dac53617d595a35cc285d2bd348baa0fbd
Soong's multi-architecture building has grown complex, with the
combination of HostOrDevice+HostType+Arch necessary to determine how to
build a variant of a module, and three separate mutators to split each
into its variations.
Combine HostOrDevice+HostType into Os, which will be Linux, Darwin,
Windows, or Android. Store Os+Arch as a single Target.
Change-Id: Iae677eff61a851b65a7192a47f2dc17c1abb4160
This lets modules declare logtags files, which will then be exported to
make using LOCAL_LOGTAGS_FILES. Make still generates event-log-tags and
the java code.
Bug: 28989759
Change-Id: Ie6be79d3ef8a17a74d42eba681a25a08b4c0e7ae
And install the tools into a more obvious location. soong_env is not
moved, since we need it to exist early, so that we can use it in
soong.bash in case there's a build failure.
Change-Id: I9bd1fa320d84d180b2cf3deb90782d380666f7a6
For the x86_arm target, which uses two 32-bit architectures, x86
libraries should be installed into ".../lib", and arm libraries should
be installed into ".../lib/arm".
This shouldn't be necessary for binaries, but non-native binaries aren't
supported in Make right now, so we can revisit this once it's necessary.
Change-Id: I4d883c85d3ef4945ff6149d9c4fc81af5023e12b
Some more common makevars methods were required in order to remove the
"-isystem" prefixes from the toolchain.IncludeFlags() value. In Make,
the -isystem is prepended at time of use, not in TARGET_C_INCLUDES
itself.
Change-Id: If07e69ddb7357d11c7dd48ab60f503d219f29de8
Colin Cross