| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
Change-Id: I3113ea6813b82c3f6b250205080030ff9c2255ec
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Looking for modules that provide a boot jar is currently
done by module name match only. This breaks when you
have multiple copies of the same module in your build
tree (eg in another device).
* Add a new mutator that marks the modules that should be used
for bootjar content.
* The logic is simple: give priority to modules that are in
exported soong namespace.
Change-Id: I2476174b892475c14a9f10b5e66b8496186f81c0
|
| |
|
|
| |
Change-Id: I5fc78a15349a77fff85b69bf2a92172149ce37fc
|
| |
|
|
| |
Change-Id: Ia7404539bf155819e15495c96238c6359b9a065c
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Soong has been outputting incorrect Android.mk entries for split apks.
soong_app_prebuilt.mk expects LOCAL_SOONG_BUILT_INSTALLED to contain a
list of <split apk path>:<on device install path> pairs whereas Soong
outputs <split apk path>:$(LOCAL_MODULE_PATH)/<split apk filename> pairs.
This also adds a missing underscore in split apk filenames.
Test: m Split
Bug: 140795853
Change-Id: I3d6506f5e12106174fbbaef34749a272b5e1b90c
|
| |
|
|
|
|
| |
Test: m nothing + diff soong mk and ninja files.
Bug: 140795853
Change-Id: I3667eac951dea7e447cf73219ff89199fca9ed63
|
| |
|
|
|
|
|
|
|
| |
This includes a few changes that make AndroidMkEntries more resemble
AndroidMkData, especially in terms of how extra entries are added.
Most importantly it can now have multiple custom functions.
Test: Soong tests
Change-Id: Ibf9102624d16d0c1c9894a2794fc7c797bb34c9a
|
| |
|
|
|
|
| |
Bug: 137218697
Test: apex_test.go
Change-Id: I55a42e1e4af60d6d7185515a380c786312b8b29b
|
| |
|
|
|
|
| |
Fixes: 136272143
Test: sh_binary_test.go
Change-Id: I2e6580286fcdf43ab20020fcd147648a3009aa9f
|
| |
|
|
|
|
| |
Fixes: 131861785
Test: sh_binary_test.go, a real sh_test with added data
Change-Id: Ic78022d2db38a530074c70823ef16773d8ba6821
|
| |
|
|
|
|
|
|
|
|
|
| |
The new system collects all Android.mk variable assignments using a map
and writes them to io.Writer. Compared to the previous system, which
directly writes all entries to buffers, this new system is more robust
and test-friendly.
Test: Built without prebuilt_etc.go change and diffed the mk output.
Test: prebuilt_etc_test.go
Change-Id: Idd28443d129ff70053295015e69328a8fa3eca47
|
| |
|
|
|
|
|
| |
Test: app_test.go
Bug: 122333215
Bug: 128610294
Change-Id: I4be98a57ffec0885258ed7d7bb2badc8b2798750
|
| |
|
|
|
|
|
| |
Bug: 128610294
Fixes: 138792623
Test: app_test.go
Change-Id: I47c80ec283ce58a0ce9b7d0af40844bd73e9d3f1
|
| |
|
|
|
|
| |
Test: app_test.go
Bug: 137218697
Change-Id: If2b20a355bb6e9e8dbcd57347c0e2c6fa041a932
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of circumventing the limitation of Prebuilt implementation by
picking a source path itself, it now uses the same mechanism as
archMutator and replaces the source path in advance so that Prebuilt
always sees the corrent source path.
Because this requires the Apk field to be a string pointer, the single
source prebuilt implementation is being updated to be reflection-based.
Test: Soong unit tests, m soong_docs, TreeHugger
Change-Id: I2304f15e32d632f74f95f0d9e9bf1f75ff3e2225
|
| |
|
|
|
|
| |
Test: androidmk_test.go, bpfix_test.go
Bug: 128610294
Change-Id: Ide183ba1e696fa0ffb4245e3288ffc47535b39af
|
| |
|
|
|
|
| |
Bug: 128610294
Test: app_test.go
Change-Id: Ie3e558bfdb40de6b0b9df95d3b373d08a4084d7b
|
| |
|
|
|
|
|
| |
This implements the uncompress-dexs macro's behavior in Soong.
Test: Converted TvSampleLeanbackLauncher + zipinfo
Change-Id: I9477aa21429d055f3f36ca90c7fd2c345c999029
|
| |
|
|
|
|
|
|
| |
This also partially consolidates shouldUncompressDex implementations.
Bug: 128610294
Test: TreeHugger
Change-Id: I7cea5a3890ddd473f63c0738a35af067455b5c4d
|
| |
|
|
|
|
|
|
| |
This is an initial version that handles the most basic cases.
Bug: 128610294
Test: app_test.go + prebuilt webview.apk
Change-Id: Ic525559aad5612987e50aa75b326b77b23acb716
|
| |
|
|
|
|
|
|
|
|
| |
log.Fatalln calls os.Exit, which skips deferred functions and
leaves the console with the cursor disabled. Use ctx.Fatalln
instead, which uses a panic to exit and calls the deferred
function to restore the cursor.
Test: DIST_DIR=/tmp/\test\ foo m dist
Change-Id: Ie92297075e37c171d5ba48848a0ddb19652b051c
|
| |
|
|
|
|
|
|
|
|
|
| |
Force soong_ui in dumpvars mode to use a dumb terminal so that it
doesn't print an empty status table. Also move the logs to files
that are prefixed with "dumpvars-" so that they don't rotate the
real build logs.
Fixes: 139307849
Test: lunch
Change-Id: If92421af07731fedc9ca42670f309e95922a2224
|
| |
|
|
|
|
|
|
|
| |
Multiple terminals have had issues with writing newlines into the
non-scrolling region, just set the cursor to the beginning of the
next line instead.
Test: m nothing in JediTerm
Change-Id: I2e434f4cc263ca13b82889a79d6a8bb48d084cb3
|
| |
|
|
|
|
|
|
| |
Don't pass 0 to setCursor or setScrollingMargins. Passing 0 hasn't
caused any known issues, but 1 is correct.
Test: m nothing
Change-Id: I8735032eb3f1d46df4c41c4ea9ef95b5fd842824
|
| |
|
|
|
|
|
|
|
| |
Keep a running map of the critical path to each edge as it
finishes. At the end of the build, find the edge with the
longest critical path and print the path to out/soong.log.
Test: critical_path_test.go
Change-Id: Ie01d26b068b768156b166bff00cdc3273e8124ca
|
| |
|
|
|
|
|
| |
Instead, fully specify build/make/target/... everywhere
Test: treehugger
Change-Id: I07ba0e9b0604919a271afd5133070616e1f404fc
|
| |
|
|
| |
Change-Id: If8fce98316be03fc9dabe27864bc4a324ab6fdb7
|
| |
|
|
| |
Change-Id: I55620e602b9e97fe7a6d36fa142ddfde77b6025b
|
| |
|
|
|
| |
Test: m, additional_manifests is now transitive
Change-Id: I3134c5e942959fe762024602fd833ec4a4312d2c
|
| |
|
|
|
|
|
|
| |
This is equivalent to 'LOCAL_FULL_LIBS_MANIFEST_FILES' in Make.
Fixes: 123374442
Test: m -j
Change-Id: Ic4b40a08f2454687c74949020598651acea3b2dd
|
| |
|
|
|
|
|
|
|
| |
Manifest merger needs --remove-tools-declarations to match
Gradle behavior.
Bug: 112607039
Test: m checkbuild
Change-Id: Id93bcaeaf03770a4acd2e1fdf44e418f55540dd3
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Don't use the merged manifest for android_library modules. We
still have to run manifest merger for android_library modules
because Make can't handle transitive dependencies, so it will
continue to merge the manifests at each library, and then merge
the manifests of direct dependencies into the final application.
Bug: 113294940
Test: m checkbuild
Change-Id: Ia8f9f910bd0a134730ddf2d542460eeddbc0a075
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Squash of:
Author: Rashed Abdel-Tawab <rashed@linux.com>
Date: Thu Aug 9 14:08:53 2018 -0700
soong: Special case Lineage SDK
* org.lineageos.platform-res.apk needs to be installed to /system/framework
* org.lineageos.platform-res needs to be a dependency for
org.lineageos.platform and org.lineageos.platform.internal
* Add other special exceptions for org.lineageos.platform-res
Change-Id: Ic617c07c086916005ea4b88f26d31c61691a45f8
Author: Sam Mortimer <sam@mortimer.me.uk>
Date: Thu Aug 30 15:33:16 2018 -0700
soong: make org.lineage.platform-res depend on framework-res
*) Allows us to build org.lineage.platform-res with no_framework_libs
true (as is done for framework-res).
*) Whilst we're here, undo a dependency loop prevention in aar.go
that we added during sdk bringup to allow our platform res to build
with no_framework_libs false.
Change-Id: Ib452a2e45112baf5d61b70b4be1ce0c01dfd84e5
Author: Luca Stefani <luca.stefani.ge1@gmail.com>
Date: Mon Feb 4 18:56:52 2019 +0100
Always link org.lineageos.platform-res for org.lineageos.platform.sdk
Test: m clean && m org.lineageos.platform.sdk
Change-Id: I58956855bd4d1157e2582103c4861e7b384b4f73
Author: Sam Mortimer <sam@mortimer.me.uk>
Date: Fri Aug 31 10:52:29 2018 -0700
soong: Allow framework to access lineage-sdk resources
*) Make framework depend on lineage-sdk resource package
*) Allows framework module to access lineage-sdk resources
via usual org.lineageos.platform.internal.R paths.
Change-Id: Ifd19d43d9308ac370ad40a499de16bf8ce204beb
Change-Id: Icc18de5dfaa83fc0a1eda6f3704f3a92e1de0764
|
| |
|
|
| |
Change-Id: Ie83688c900a241728b1153cc8ac2aeccc2cb31f6
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Used by vendor/lineage generated kernel header module.
Partial pick from:
Author: Sam Mortimer <sam@mortimer.me.uk>
Date: Fri Aug 17 11:25:08 2018 -0700
soong: Add java sources overlay support
Change-Id: I94143febb0a8afa6a165364d36a40d5120a4e7bc
Change-Id: I415af71458f2a7be8e256cb3c548994f09c5bebf
|
| |
|
|
| |
Change-Id: I0e5395ac70220f1d3a1c87c6112e33f84f526fea
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Includes the following commit:
Author: Simon Shields <simon@lineageos.org>
Date: Mon Sep 4 21:09:44 2017 +1000
soong: explicitly specify name of Lineage variables struct
Change-Id: Ic754b8a13658ddc40620f985581bd1754b780b15
Change-Id: If127c9da6b57340ae57e0cf725151cdcc3c82ef3
|
| |
|
|
|
|
|
|
|
|
| |
*) Add option --copy-all-files that copies all generated output.
*) When enabled, do not error when unexpected output files are found.
*) For use by lineage_generator module.
Change-Id: Ia3cd2216807f6165b23a97a001db7303f15d33e2
|
| |
|
|
| |
Change-Id: I9af8579f4acf5043375650ec4f7fadb66c0eb15e
|
| |
|
|
|
|
|
|
| |
-Allow folks to see easily if the term session they are in has
WITH_SU set to true
-Also add a dumpvar rule for WITH_GMS for those that have it
Change-Id: Ib310d0016bb9eb16960f794720321b706b6fd8b0
|
| |
|
|
| |
Change-Id: I24b29442f041aec451a173db404f542060ac67d2
|
| |
|
|
|
|
|
|
|
|
| |
Don't enforce namespaces on Blueprint modules like
bootstrap_go_package, their dependencies are handled before
namespaces are initialized in namespaceMutator.
Fixes: 135246048
Test: TestDependingOnBlueprintModuleInRootNamespace
Change-Id: I7cf1c26bb8512eed59d6b4eb42a49f7080ffa281
|
| |
|
|
|
| |
Test: treehugger
Change-Id: Icdbe62bcb7bc4717228c5d974962b939d8eafee6
|
| |
|
|
|
|
|
|
|
|
| |
mm is now doing the same thing as mma. Now print a deprecation message
for other tools that may be setting ONE_SHOT_MAKEFILE directly (or users
that are still using a very old envsetup.sh).
Test: mm # does not print
Test: ONE_SHOT_MAKEFILE=bionic m
Change-Id: I98221657024d824a61c0d721d5c7dd1a2ea2fb1e
|
| |
|
|
|
|
|
|
|
|
|
| |
If OUT_DIR has been specified through the command line arguments,
it is sanitized first and NewConfig sets the OUT_DIR in the build
environment. When calling OutDir from Config instance, the out
was being sanitized again which is unnecessary.
Bug: b/118730755
Test: Ran unit test cases, "lunch" followed by "m" command.
Change-Id: Ieb4ffa1026c802c76c7369e8cdb3923a4c70f4b3
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, any of the build commands (mma, mmma, etc...) would traverse
up the directory path until an Android build file has been found. That
directory path would eventually become the MODULES-IN-<dir name> build
target. The directory where the m* command was executed may contain a
build file in a sub directory. For that case, the build target is the
MODULES-IN-<dir name> where dir is the directory of the executed m*
command since the target has been constructed by the core makefile.
Bug: b/118730755
Test: Wrote and ran unit test cases. Ran "lunch", "m". Went to a
directory in AOSP that had build files in sub directories and
executed mma. Went to a directory that had only build files in
parent directories and executed mma.
Change-Id: I9458cbc4bf0fc25bfb0848e96e92d261442d5cb0
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For mmma and mmm, the findBuildFile function in config.go is invoked every time
for specified directories and directories with targets. For directories with
targets, an Android build file must exist in the directory where mmma and mmm
was invoked. There is no need to invoke findBuildFile function as a
simple check of the build file exists in the specified directory.
This is also refactoring the code for b/118730755
Bug: b/118730755
Test: Executed unit test cases through Intellij and executed mmma
command: "mmma external/protobuf:aprotoc external/bzip2".
Change-Id: I5428e3a3c36f77ff740617564b7853705521f29f
|
| |
|
|
|
|
|
|
|
| |
This is the first part of deprecating the mm command. mm will simply
be redirected to mma by building the modules and their dependencies.
Bug: b/135187558
Test: Ran mm in external/protobuf directory.
Change-Id: Idad40f89e3768aeb459172f632a44fd7b0a201bc
|
| |
|
|
|
|
|
|
|
|
| |
This is the first part of deprecating the mmm command. mmm will simply
be redirected to mmma by building the list of modules with their
dependencies.
Bug: b/135187558
Test: Ran mmm external/libusb:libusb.
Change-Id: I50cf0d6a07197556ab6130e612561c6d5a3fb31e
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The build directory passed in to soong_ui can contain a
symbolic link. soong_ui was not evaluating the build
directory to retrieve the true path, hence failing to
execute the internal soong.
Fixes: b/135995632
Test: Unit test case and executed the scenario in the bug
description.
Change-Id: I5779c6aa3f3183810437dbe2b2d4e40acbafb205
|