The existing NormalizePathForTesting function does not handle make
install paths very well, as it returns a relative path with a leading
"../" which is very confusing. It also does not clearly differentiate
between the different paths.
These functions return paths that are basically what are seen in a
normal developer build, i.e.
* <source path>
* out/soong/<soong output path>
* out/<make output path>
That makes tests that use them easier to understand.
Follow up changes will clean up the existing usages of the
Normalize... functions.
Bug: 182885307
Test: m nothing
Change-Id: I17ddc996bef5bbbf4a62da8334ea6ce29e306109
Prepare for using Config when adding singletons by passing
Config to NewTestContext and NewContext instead of to ctx.Register.
This will enable a followup change to store SingletonMakeVarsProviders
registered on the Context in the Config, which is necessary to run
multiple tests in parallel without data races.
Test: all soong tests
Change-Id: Id229629a4e42ff4487d317241673837726c075fc
Add vendor_ramdisk_available and vendor_ramdisk attribute to
various rules. When a vendor_ramdisk variant of a module is
generated, it is installed to $OUT/vendor-ramdisk.
It is similar to a ramdisk image.
Test: m nothing -j
Change-Id: Ib2d16459f3094dbe21c3bdb7c016cb4b2bf62765
This change introduces the concept of partition dir for InstallPaths.
It's the path to the partition where the InstallPath is rooted at. For
example, it's out/soong/target/product/<device>/<partitoon> for paths
created for device modules. For host modules, it is defined as
out/soong/host/<host_os>-<host_arch>.
The partition dir is obtained using the new PartitionDir() function.
Another change is that a freshly created InstallPath (usually via
PathForModuleInstall) is the result of joining PartitionDir() and the
remaining path elements. For example, PathForModuleInstall(ctx, "foo",
"bar").Rel() now returns "foo/bar". Previously, that call returned the
relative path from config.buildDir() ("out/soong"). This change is in
line with the behavior of other path-creating functions like
PathForModuleSrc where Rel() returns the path relative to the
contextually determined path like the module source directory.
Notice that the Join() call to InstallPath doesn't change
PartitionDir(), while does change the result of Rel().
p := PathForModuleInstall(ctx, "foo", "bar")
p.PartitionDir() is out/soong/host/linux-x86
p.Rel() is foo/bar
q := p.Join(ctx, "baz")
q.PartitionDir() is still out/soong/host/linux-x86
q.Rel() now returns baz
Bug: N/A
Test: m nothing
Change-Id: I916bb1c782a4bfe0fbd4854e349cd2a2a42f56b6
This reverts commit 323dc60712.
Reason for revert: Possible cause of test instability
Bug: 170513220
Test: soong tests
Change-Id: Iee168e9fbb4210569e6cffcc23e60d111403abb8
Putting t.Parallel() in each test makes them run in parallel.
Additional t.Parallel() could be added to each subtest, although
that requires making a local copy of the loop variable for
table driven tests.
Test: m checkbuild
Change-Id: I5d9869ead441093f4d7c5757f2447385333a95a4
This change fixes some places where host arch is assumed to be x86 to
form a path element like linux-x86. In preparation for non-x86 host
targets, the host arch part is derived from the context.
In addition, InstallForceOS() is refactored so that it can override not
only OsType, but also ArchType. Without this, the paths for the
robolectic modules will be changed from linux-x86 to linux-common, which
breaks several other places where the old paths are expected.
Bug: 134795810
Test: m
Change-Id: Ib38c715948ae546e55021ece82bac1d82e9e5da0
"excludes" should be applied both SourceFileProducer and
OutputFileProducer. But we've been missing the latter.
(see android/paths_test.go)
There's one more to note. Previously, the exclusion was done by
modifying slice directly, which caused some problem(b/76179848).
The fix was returning a copy from Srcs() method. But there're still many
implementors which just return internal slices directly.
This change creates a new array when it needs to exclude something from
returned slice.
Bug: n/a
Test: m (soong test added)
Change-Id: I1ba91e490cb55fc243020681667278cdc75f59df
FirstUniquePaths is called on some long lists where the O(n^2)
behavior is problematic. Use a map-based implementation for
longer lists.
Test: TestFirstUniqueStrings
Change-Id: I7181aba869e5ccc0f99c2fa7b8f03839f06e4307
Robolectric tests compile against device modules but are installed
and run as host modules. Allow a module to override its install
OS.
Test: TestPathForModuleInstall
Change-Id: Icf37bb3d4cc1222a9b079602c6a5fdb8b51c86ed
This relands I12a0f907753fefd1997ab8b4ea2ac331234093cf along with
a fix to blueprint for absolute paths.
Store the current working directory and then change to the root
directory so that all file accesses must go through helpers in
the android package that properly track dependencies.
Change-Id: I24ac485677aa102eec1a2521d16820da6ee1ae77
Fixes: 146437378
Test: m checkbuild
Test: m OUT_DIR=/tmp/out nothing
Store the current working directory and then change to the root
directory so that all file accesses must go through helpers in
the android package that properly track dependencies.
Fixes: 146437378
Test: m checkbuild
Change-Id: I12a0f907753fefd1997ab8b4ea2ac331234093cf
Make LoadHookContext embed a new EarlyModuleContext instead of
BaseModuleContext to reduce its API surface in preparation for
moving it to run during parsing instead of mutators.
Test: m checkbuild
Change-Id: I1cd3ff3b636e7e24991a9184d7521903473e505a
The filesystem object was available through ModuleContext.Fs(), but
gives too much access to the filesystem without enforicing correct
dependencies. In order to support sandboxing the soong_build
process move the filesystem into the Config. The next change will
make it private.
Bug: 146437378
Test: all Soong tests
Change-Id: I5d3ae9108f120fd335b21efd612aefa078378813
Run the imageMutator between osMutator and archMutator so that
different arch variants can be set for the different partitions.
Bug: 142286466
Test: m checkbuild
Change-Id: I65d05714b75aa462bf9816da60fdc2deda4de593
Merged-In: I65d05714b75aa462bf9816da60fdc2deda4de593
(cherry picked from commit 9c8f687584)
Avoid having to pass ModuleFactoryAdaptor to every call to
RegisterModuleType in a test by wrapping RegisterModuleType.
Test: all soong tests
Change-Id: If8847d16487de0479cc3020b728256922b3cadba
If InstallInRoot() returns true the module will be installed to
$OUT/root or $OUT/recovery/root.
Bug: 141877526
Test: m checkbuild
Test: no change to build.ninja or Android-${TARGET_PRODUCT}.mk
Test: TestPathForModuleInstall
Change-Id: Id6e435c6019f11eeb5806528fd464dbf220b88d9
Allow modules to mark themselves as InstallBypassMake, which will
cause android.PathForModuleInstall to return a path in $OUT_DIR
instead of $OUT_DIR/soong. This can be used for modules that
can handle installation to the final location on their own. The
main blocker for most modules is support for the "required" property,
which requires adding dependencies on the installed location of
other modules.
Bug: 122332855
Test: m checkbuild
Change-Id: I85238d937ff30335167d4b3fec79bbefc734b5e1
There is no need to create a separate buildDir for each test
file, use TestMain to create a global one for the package.
Test: all soong tests
Change-Id: I435ee7aa88b7e0bb8ccc1ba79f82833a7accf3e9
AddMissingDependencies is needed in BaseModuleContext in order
to allow PathsForModuleSrc and friends to be called in mutators,
It's not a simple move, as it currently writes to a module
field that would be lost if any mutators cloned new variants
by calling CreateVariations. Reimplement it using a mutated
property instead.
Test: m checkbuild
Change-Id: I851125065e4c5302b552773dae4640426c62965e
blueprint.BaseModuleContext is the set of methods available to all
module-specific calls (GenerateBuildActions or mutators). The
android package split the same functionality across baseContext (nee
androidBaseContext), BaseModuleContext, and BaseContext.
Consolidate all of them into android.BaseModuleContext.
Test: m checkbuild
Change-Id: I2d7f5c56fd4424032cb93edff6dc730ff33e4f1e
There are cases where a module needs to refer to an intermediate
output of another module instead of its final output. For example,
a module may want to use the .jar containing .class files from
another module whose final output is a .jar containing classes.dex
files. Support a new ":module{.tag}" format in any property that
is annotated with `android:"path"`, which will query the target
module for its ".tag" output(s).
Test: path_properties_test.go, paths_test.go
Test: no unexpected changes in build.ninja
Change-Id: Icd3c9b0d83ff125771767c04046fcffb9fc3f65a
Set the location label for missing srcs and tools to avoid
nonsensical errors when parsing the command.
Test: genrule_test.go
Test: paths_test.go
Test: unbundled branch with missing framework-res module needed by robolectric genrule
Change-Id: I9c1f1cd82a80f048c0e903b8e93910b1ae34b0b1
Move the logic from ctx.ExpandSources into android.PathsForModuleSrc
and ctx.ExpandSource into android.PathForModuleSrc, and deprecate
them. When combined with the pathDepsMutator this will let all
properties that take source paths also take filegroups or genrule
outputs, as long as they are tagged with `android:"path"`.
Test: All soong tests
Change-Id: I01625e76b5da19240e9649bf26a014eeeafcab8f
ModuleSrcPath was designed as a type that ensured that modules only
referenced sources inside the directory that contained the Android.bp
file. In practice they don't work very well, because allowing
filegroups and genrules as inputs to any module that takes a source
path means that the path might end up being to a file in another
source directory or to a generated file in the output directory.
Remove ModuleSrcPath, replacing it with SourcePath in the places
that need to explicitly refer to a path in the source tree, or
Path where it may be a source path or a generated path.
Make PathForModuleSrc return a Path instead of a SourcePath in
preparation for consolidation with ctx.ExpandSources, which will
make it possibly return paths to generated files.
Test: All soong tests
Change-Id: I973a78470ed14307eea5f6d0cc93942775a65715
Copying p.rel doesn't work, as rel needs to match the end of the
value in path. Apply the same transformation to p.rel as p.path.
Test: paths_test.go
Change-Id: I42d676c6c4fc18d9852c1a73f25e5a791d7553d0
Add PathsForOutput to convert multiple strings into WritablePaths.
Add OutputPath.InSameDir to build a new OutputPath pointing to a
file in the same directory as an existing OutputPath.
Add WritablePathForTesting and WritablePathsForTesting that mirror
PathForTesting and PathsForTesting but return WritablePaths.
Add PathContextForTesting to return a minimal PathContext
implementation.
Test: paths_test.go
Change-Id: I9708eb164b273514a96dae0a260ef9a963fb9bcf
Add a helper method that returns a new OutputPath with the
extension replaced.
Test: paths_test.go
Change-Id: I7f93b3475031a19b53652022740488373a1b7d97
filepath.Abs is surprisingly expensive, it calls os.Getwd every
time, which involves multiple syscalls, a lock, and and allocations.
Use IsAbs and prefix matching instead.
Test: paths_test.go
Change-Id: Ia6cf34d6bef24c694702af1e7a6ff08ffd2d822b
Port the dexpreopt logic from Make to the dexpreopt package in Soong,
and use it to dexpreopt Soong modules. The same package is also
compiled into the dexpreopt_gen binary to generate dexpreopt scripts
for Make modules.
This relands Ib67e2febf9ed921f06e8a86b9ec945c80dff35eb and
I462182638bd57b1367b5bfb0718e975c11ae66f7, along with multiple fixes
to depsfile generation in dexpreopt_gen that caused .odex files for
modules in defined make to be missing dependencies on boot.art, and
a fix to not dexpreopt and strip tests.
Bug: 119412419
Bug: 120273280
Test: no differences to dexpreopt outputs on aosp_sailfish system/,
only expected changes to dexpreopt outputs on system_other
(.vdex files for privileged Soong modules no longer incorrectly
contain .dex contents).
Test: OUT_DIR=$PWD/out m
Test: NINJA_ARGS="-t deps out/target/product/sailfish/obj/APPS/Contacts_intermediates/dexpreopt.zip" m
Change-Id: I6bb2c971cee65d2338839753aa0d84939f335b1b
MaybeRel was failing with an error when comparing an absolute out
path to a relative source path. Make it return false instead, and
add tests.
Bug: 119412419
Test: paths_test.go
Change-Id: I462182638bd57b1367b5bfb0718e975c11ae66f7
Attempting to reduce the number of different spellings we have for
"product services" partition in the codebase.
Bug: 112431447
Test: m
Change-Id: I0a393a1d625e7ea3217d28735a4db709bce32395
Merged-In: I0a393a1d625e7ea3217d28735a4db709bce32395
This is an adaptation of Icc4f8c16bc389fe20db680849f311d02df1299c3, to
support modules that are installed on the /product-services partition.
Bug: 80741439
Test: m -j both with and without enabling the new partition
Change-Id: I72b335ad38baff5848cd3da7489343f8cf98ff16
`recovery: true` installs a module to the recovery partition.
`recovery_available: true` makes a module to be available to other
`recovery:true` or `recovery_available: true` modules.
These to are very similar to vendor, vendor_available properties, except
for the target partition.
Bug: 67916654
Bug: 64960723
Test: m -j, toybox_recovery is installed to the recovery/root/sbin
Change-Id: Iaebe0593de16c69fa70de251a61f4d018a251509
LOCAL_RESOURCE_DIRS puts the highest priority overlay first, but
aapt2 expects the highest priority overlay last. Soong stores the
list in aapt2 order (low to high priority), but that means when it
exports to Make as LOCAL_SOONG_RRO_DIRS, which goes to
build_rro_package.mk and then package_internal.mk, it gets reversed
again and comes out backwards.
Bug: 78032566
Test: m checkbuild
Change-Id: If72bf929fbf1d126f9051a2f21ec1eb4e3030e6e
The next patch will need to more complicated custom error handling,
so make validatePath return an error and let the caller handle it.
Test: paths_test.go
Change-Id: I4fe11c3f319303d779596709f4819e828b5bdb9b
