This CL refactors the code related to ManifestFixer parameters.
The required parameters android.ModuleContext, manifest android.Path are
passed separately as the parameters and the optional parameters are
kept as part of the ManifestFixerParams struct.
By default, the member variable of struct have the zero (nil, false,
empty string) values. Hence, it is only required to pass the
parameters of interest at the time of function call to
ManifestFixer.
Manual testing done to check the working of the code.
Test: m nothing && m test_com.android.sdkext
Test: manually tested the generation of AndroidManifest in the out
directory with the testOnly attribute
Test: atest manifest_fixer_test --host
To test the existing unittests are not breaking.
Change-Id: I20cb6c06c57f8fe7811050288bcb03945dc0425b
If the build file contains the apex_test module, but the apex_test
module does not contain the AndroindManifest file, then create the
AndroidManifest file.
In such case, the apexer tool is already generating an AndroidManifest
file. In order to handle the testOnly attribute for apex_test modules, a
--test_only flag is appended to the opt flags.
The apexer tool reads the opt flags and if the --test_only flag is
present then it generate the AndroidManifest file with testOnly
attribute.
Bug: 213310150
Test: m nothing && m test_com.android.sdkext
Test: manually checked the generation of AndroidManifest file in the
unsigned zip file mentioned in the description.
This command allows to read the attribute of the binary xml file -
aapt2 dump xmltree test_com.android.sdkext.apex.unsigned --file
AndroidManifest.xml
Verified the presence of testOnly flag in the output.
Change-Id: Ic47378428b2dba51d73e75d912546c2374f68d57
In case there are two vendor apexes(one with "use_vndk_as_stable:true",
and the other with "use_vndk_as_stable:false") a VNDK lib used by both
will have "APEX" variant and the former APEX will use "apex" variation.
For example,
apex1(use_vndk_as_stable) -> foo -> libvndk
apex2 -> bar -> libvndk
Since foo, bar and libvndk are mutated into two APEX variations("",
"apex10000"), foo will use the apex variation of libvndk.
To fix this, VNDK libs can use "unique" APEX variations. Then, in the
above example, foo will have "myapex1" variation and libvndk will have
two APEX variations("" and "apex2"). So foo will link to ""(non-APEX)
variation as fallback.
Bug: 216847402
Test: m nothing (soong tests)
Change-Id: I116932860ef79e22dc338a58b251e3ca693ab4f3
Even though a vendor APEX sets use_vndk_as_stable:true it was possible
to include a VNDK lib by directly depending on it with
native_shared_libs.
But it's contradictory to have a VNDK lib while declaring not to include
VNDK libs. It was missing since pruning dependencies on VNDK libs was
done only for transitive deps.
Added a check to reject this.
Bug: 216847402
Test: m nothing(running soong tests)
Change-Id: I8d79a434b1bfe8e563cf8968fa76830b0e582f66
If the build file contains the apex_test module, add the
testOnly attribute to the application element of the
corresponding AndroidManifest file and set its value to true.
If the testOnly attribute is already present and has value
false, then do nothing.
Tests added in manifest_fixer_test.py to check if the updated
AndroidManifest file has the testOnly attribute set to true or not.
Bug: 213310150
Test: atest --host manifest_fixer_test
Test: m nothing
Test: manually checked the AndroidManifest file generated
Change-Id: I36247dbe0261c342d451a4422c314fd8fe0c2369
This change is similar to aosp/1947127, but for prebuilts.
After this change, if `bootImageConfig.installDirOnDevice` is set to a
path outside of the APEX, the build system will build a boot image from
the dex files and the profile extracted from the prebuilt APEX.
Otherwise, it keeps the current behavior: extracting the boot image from
the prebuilt APEX.
This is a no-op change. Current behavior is not affected.
Bug: 211973309
Test: m nothing
Test: -
On internal master:
1. Patch aosp/1947128.
2. Patch ag/16743847 and ag/16746804.
3. m SOONG_CONFIG_art_module_source_build=false com.google.android.art
4. See the boot image being installed in `/system/framework/<arch>`.
Change-Id: I24ca525309fecaf3ab7a67960fbf118cd00ecd1d
Bug: 214466457
Bug: 207551677
Test: b build //build/bazel/examples/apex/minimal:build.bazel.examples.apex.minimal
Test: b test //build/bazel/tests/apex:build.bazel.examples.apex.minimal_apex
Test: b test //build/bazel/tests/apex:build.bazel.examples.apex.minimal_capex
Change-Id: I6bf12c1b0c52d4486968bb430a67a3c3110766db
This is needed to allow the Android Go versions of modules
(com.google.android.go.*) to select a different compression toggle than
their non-Go (com.google.android.*) conterparts. go/go-updatability
Bug: 203024418
Test: Preloaded go variants on wembley, booted Android.
Change-Id: Ic96aff5fafb65fbd08e8a69d47c994532e27819e
Merged-In: Ic96aff5fafb65fbd08e8a69d47c994532e27819e
(cherry picked from commit 2c4a96375a)
Discourage jarjaring code where there are alternatives with
better system health implications.
Test: m
Bug: 215233995
Change-Id: I1b076d00e1ad6aa32b41da6bda1033978b5e829d
Currently, the bpf module netd.o (source system/netd/bpf_progs/netd.c)
will be built to /system/etc/bpf/netd.o. In Android T, it will be moved
to mainline module com.android.tethering.
The expected behavior is:
- In T device, it uses the netd.o in mainline module.
- In pre-T devices, it uses the original netd, built from platform.
However, netd.o will be double loaded if the tethering module is
installed in Pre-T devices. Because:
1. bpf in apex is packed into /apex/MAINLINE_MODULE/etc/bpf/
2. bpf in platform is packed into /system/etc/bpf/
3. bpfloader in pre-T loads ANY bpf modules under
/apex/com.android.tethering/etc/bpf/ and /system/etc/bpf/.
We can't change the behavior of bpfloader in pre-T devices. We can't
delete the /system/etc/bpf/netd.o from pre-T devices. Both of them are
not mainline modules. So the mainlined netd.o needs to be packed into a
folder other than /apex/com.android.tethering/etc/bpf/ or
/system/etc/bpf/.
This commit adds a tag 'sub_dir' for bpf module. The installation path
of bpf modules will be:
- /system/etc/bpf/SUB_DIR/ (for platform code)
- /apex/MAINLINE_MODULE/etc/bpf/SUB_DIR/ (for mainline module)
Bug: 202086915
Test: add test in apex_test.go and build
Change-Id: Icc6619768ab006de9f86620a7df1bb2853eaba13
After this change, `bootImageConfig.installDirOnDevice` can be set to a
path outside of the APEX, in which case, the boot image will not be
installed in the APEX. Instead, it will be installed to the given path
by Make.
This is a no-op change. Current behavior is not affected.
Bug: 211973309
Test: m nothing
Test: -
1. m com.android.art
2. See the boot image still being installed in the ART APEX.
Test: -
1. Change `installDirOnDevice` of the ART boot image config to
`system/framework`.
2. See the boot image being installed in `/system/framework/<arch>`.
Change-Id: Ib13b17cc9e94dc5754c9b51b04df3307323b8783
This change adds support for dexpreopting standalone system server
jars from prebuilts.
Bug: 203198541
Test: -
1. Add a standalone system server jar (e.g., by patching
aosp/1906158)
2. Build and drop a module SDK and an APEX.
3. Build a system image from prebuilts.
4. See the odex and vdex files generated in
$ANDROID_PRODUCT_OUT/system/framework/oat/
Change-Id: I8f4eaed10a1053cd560b8583efa12dc495f58db1
The property is used to customize uid/gid/mode/capabilities of files in
an APEX.
Bug: 209971551
Test: m
Change-Id: I484e46ff819a5266c1e8046dae337e18ef3fefea
... in preparation for adding the support for custom canned fs config
Bug: 209971551
Test: m nothing
Change-Id: I7f2576ff99c65bdb6c9ce4ace61bc783eea2f0d4
App compact impact is hard to judge (see http://b/124106384#comment9
ff), so this tries it early in the dessert cycle.
This is also a workaround for b/209867137 where the symlink install
rules interact badly with the artifact_path_requirements.mk check.
Test: m installclean && m droid
verify that $(PRODUCT_DIR)/system/bin doesn't have dalvikvm and
dex2oat symlinks
Bug: 124106384
Bug: 209867137
Change-Id: I044fb40f656beeb7c6e61f1d418fa13f56714a70
Each conversion required defining a separate mutator, which will each
operate on _all_ modules and requires each to repeat checks whether the
mutator should operator. Instead, we introduce a single mutator and
modules can define a ConvertWithBp2build to implement bp2build
conversion for that module.
Test: bp2build.sh
Bug: 183079158
Change-Id: I99d4b51f441c2903879092c5b56313d606d4338d
The ART AOT exemption only applies to Q/R/S, so module jars that have
min_sdk T+ do not need to follow the module package restriction, even if
they are part of a Q/R/S module (but not loaded on Q/R/S).
Relax the restriction to only apply to jars that have min_sdk before T.
Bug: 208773835
Test: m (runs apex tests)
Change-Id: I2c3ad8984ca05ad763bf6162bd478f93ab4ee650
apex_set modules failed to set prebuiltCommon.installedFile, which
caused LOCAL_SOONG_INSTALLED_MODULE to be empty and base_rules.mk
to create a duplicate install rule. Set it to the path to the
apex file installed by Soong.
Bug: 204136549
Bug: 209867137
Test: m SOONG_CONFIG_art_module_source_build=false
Change-Id: Ia7fec09598823343242ebd44f1732e6bba21d027
Soong has enough information to build the license metadata files
without resorting to the fixups required in Make.
Bug: 207445310
Test: m checkbuild
Change-Id: I8e74108376162b8fdb87ba098ebe94350aa1f7c4
Inidividual boot or system server jars may have higher min_sdk_version
than the contianing apex, since the runtime respects the values of
min/max_sdk_version; e.g. runtime would not load a boot jar with
higher min_sdk_version. This allows shipping new boot jars via apexes
that target older platforms.
Bug: 190818041
Test: presubmit
Change-Id: I08ec0b4463a17bc8265b948fe09da55eb4e52ac3
apex.binaries accepts a list of module names, which should be resolved to
their fully qualified labels. Bazel treats string_list and label_list
attributes differently, most notably with the latter adding dependency edges.
Test: apex_conversion_test.go
Bug: 209743852
Change-Id: Iafdc5c728e8658cce0e99d42f32e7bb6fe2f3168
This is configured from Make by setting up Darwin+Arm64 as a HOST_CROSS
target (which is largely true, as binaries can't be executed on X86_64
machines). On the Soong side, it's a bit blurier, as we don't current
have any other users that are the same OS but not natively executable
(Linux/musl is executable, Windows/x86 is a different OS).
Instead of requiring cc_binary/etc to become multi-architecture-aware
and using something like common_first/MultiTarget, this defaults all
non-multi-architecture-aware modules to be built with both
architectures. It then adds a dependency with the
DarwinUniversalVariantTag so that supporting modules can get the outputs
of the other variant.
Cc uses that dependency tag to run lipo on shared libraries and binaries
so that the output of the x86_64 variant is actually a fat binary
including both architectures.
Bug: 203607969
Test: build sdk-repo targets on a Mac
Change-Id: Icbddb0a177c0ba19d3e0d11f8cf568e0d1ea3245
Without these errors, the last encountered deapexer would silently be
used. However with PRODUCT_INSTALL_APEXES there will only be deapexer
for installable APEXes, and we can assume only a single installable
APEX exists for each APEX variant, so make it an error if it isn't the
case.
Corresponding to http://ag/15156931.
Test: m nothing SOONG_CONFIG_art_module_source_build=true
Test: m nothing SOONG_CONFIG_art_module_source_build=false
Test: m nothing SOONG_CONFIG_art_module_source_build=false
with installable:true for the com.android.art.debug prebuilt APEX -
check that it fails with an "ambiguous duplicate deapexer" error
Bug: 192006406
Bug: 192542393
Change-Id: I44566fd26b12f82a8a67fe4a69e56303460756d0
