The default build-id is generated using timestamp. Pass -Brepro to
avoid using timestamp for deterministic build.
Bug: 153462962
Test: build fastboot.exe twice, got same shasum
Change-Id: I38fe993eec23c60bfcf1b76188774bfe06839fa4
Merged-In: I38fe993eec23c60bfcf1b76188774bfe06839fa4
(cherry picked from commit a3c22e7ee2)
Exempt-From-Owner-Approval: backport
An attempt to reference one of the prebuilt modules directly from an
Android.bp file highlighted a bug. The prebuilt_apis module used a
mutator to create filegroup and java_import modules for a set of api
versions. That mutator ran after the "prebuilts" mutator which handled
the renaming of prebuilt modules when the matching source module was
not present. That meant that the only way to reference the module was
by explicitly adding the prefix "prebuilt_".
This change fixed that bug by creating the modules in a load hook rather
that a mutator. This ensures that the prebuilt modules are present well
before the "prebuilts" mutator is run. Removing the mutator also removes
an unnecessary traversal of the whole dependency tree.
It also updated the documentation to explain that it creates
java_import modules as well as the filegroup modules.
Bug: 153649372
Test: m droid
Change-Id: Id44dce8ca9968ae903345df6ef1c4f1be9cb76c5
Remote execution and other tools can be confused by references to
build_number.txt without a dependency. Add an order-only dependency,
which maintains the current behavior.
Test: BUILD_NUMBER=1 && m aapt && aapt version # shows 1
BUILD_NUMBER=2 && m aapt && aapt version # shows 1
rm out/soong/.intermediates/frameworks/base/tools/aapt/aapt/linux_glibc_x86_64/aapt
BUILD_NUMBER=2 && m aapt && aapt version # shows 2
Change-Id: Icfa98d6840b1dc2e273ba29c33011635d1cf93b1
Bug: 153071808
Merged-In: I1609a790dd4d0a03c8308b6e552622fe33fa2499
This is to guard against the potential situation when someone adds
updatable modules to the list of boot jars by mistake.
Test: aosp_walleye-userdebug builds.
Test: Manually break the checks and observe the errors:
- move updatable module 'conscrypt' from
PRODUCT_UPDATABLE_BOOT_JARS to ART_APEX_JARS:
internal error: module 'conscrypt' from updatable apex 'com.android.conscrypt' is not allowed in the ART boot image
- add updatable module 'conscrypt' to ART_APEX_JARS
(but do not remove it from PRODUCT_UPDATABLE_BOOT_JARS):
error: A jar in PRODUCT_UPDATABLE_BOOT_JARS must not be in PRODUCT_BOOT_JARS, but conscrypt is.
- move updatable module 'framework-tethering' from
PRODUCT_UPDATABLE_BOOT_JARS to PRODUCT_BOOT_JARS:
internal error: module 'framework-tethering' from updatable apex 'com.android.tethering' is not allowed in the framework boot image
- add non-updatable (in AOSP) module 'android.net.ipsec.ike'
to PRODUCT_BOOT_JARS:
internal error: failed to find a dex jar path for module 'com.android.ipsec.ike', note that some jars may be filtered out by module constraints
Bug: 147579140
Exempt-From-Owner-Approval: cherry-pick, approved in AOSP.
Change-Id: I25ca2f52530fcfa1f9823b2cfa3485db9c0d0db1
Merged-In: I25ca2f52530fcfa1f9823b2cfa3485db9c0d0db1
(cherry picked from commit b28cc3758c)
Modules with compile_multilib property set to "both" cause builds to
fail due to duplicate Make recipes. Fix it by outputting reports only
for the first target for now.
Bug: 153461674
Test: Built cts with the coverage flags set.
Change-Id: Ibb8da6e9c112a857695cd7095a62c910cddfbb94
It has been wrong to split ALL_VERSIONS into exclusive two sets of
before/after TARGET_PLATFORM_VERSION.
And PLATFORM_VERSION_ALL_CODENAMES supports all *active* list of
non-finalized codenames.
Bug: 152960049
Test: m
Change-Id: I78ca88758998e440bea72ba2d56d90eea3ec99ae
Golang's FileInfo() implementation recognizes FAT filesystem directories
correctly, but doesn't actually add executable bits, which essentially
makes them inaccessible. Use 0777 when we know we're dealing with
directories.
Fixes: 153207404
Test: manually tested with an aar file with FAT permission bits.
Change-Id: Ie9f5320bc44d24b2f9d54d4036bb1747456655bb
Using a lot of boolean variables can become very verbose without adding
really any new information:
variables: ["a", "b", "c"],
}
soong_config_bool_variable {
name: "a",
}
soong_config_bool_variable {
name: "b",
}
soong_config_bool_variable {
name: "c",
}
Now turns into:
bool_variables: ["a", "b", "c"],
}
Bug: 153161144
Test: built-in tests
Change-Id: If5455a38433431c7ecbce1e5b32cfbb47f42602a
Merged-In: If5455a38433431c7ecbce1e5b32cfbb47f42602a
(cherry picked from commit 2b8b89cfa2)
This causes the compiler to emit some additional debug infomation that
will be used for sampling PGO. These debug infomation will get stripped
so it only affects intermediate files.
Test: build
Bug: 79161490
Bug: 153039105
Change-Id: Ie4d1d5ffbd311ba6e268cb94a618f5272be246ef
Merged-In: Ie4d1d5ffbd311ba6e268cb94a618f5272be246ef
(cherry picked from commit ceb5b76c91)
Make `permitted_packages` mandatory for updatable boot class
path jars. Collect the permitted_packages from those jars to
a file installed as /system/etc/updatable_bcp_packages.txt .
(cherry picked from commit 205e6c2a15)
Test: aosp_taimen-userdebug boots.
Test: adb shell cat /system/etc/updatable-bcp-packages
Test: Manual, remove permitted_packages from framework-tethering,
build fails.
Bug: 151314205
Merged-In: I21def97ace9081e707910d449943c683189f16cf
Change-Id: I68486f0d8d3368636e1a5324321bd0106fbe241a
We plan to add boot image variants for host tests.
Distinguishing the variants by arch does not work,
since both host and device can have the same arch.
(cherry picked from commit c177ebeca9)
Test: m
Bug: 152037801
Merged-In: Iea73c77367affb074f97a0fc318389417ce537da
Change-Id: I8b95882bb00cd2fd1f6cfd8a2784f9ebb957eec6
It looks like sometime in late February our Mac builds started taking
~10 minutes longer than before. On my local workstation the Soong tests
were taking >25 minutes before completing (likely because I don't have
the older SDKs installed, and we iterate from older to newer to find the
oldest installed SDK).
Most of this time was spend running the `xcrun` tools to interrogate the
system about which Mac SDKs are installed and where the tools are. This
will never change during any build or test, so wrap it all in a
sync.Once so that we only ever call them once.
And remove the macSdkPath variable, which has been unused for years and
no longer works (as we don't allow the use of xcode-select during the
build).
Bug: 153010389
Test: prebuilts/build-tools/build-prebuilts.sh on a Mac
Change-Id: I39b2d49739e628e4c11bec4805b25039115d2fd0
Merged-In: I39b2d49739e628e4c11bec4805b25039115d2fd0
(cherry picked from commit 6ba5367a70)
The partition tag helps merge_target_files.py determine
which apexkeys.txt to select from the framework partial
target files and which to select from the vendor partial
target files. The partition tag is the pysical partition
name, for example, a system_ext module on a device where
system_ext is at system/system_ext has a system
partition tag.
Bug: 138942268
Change-Id: Ia07887b34f1aa77dae94ef23610dfef83c1a5849
Merged-In: Ia07887b34f1aa77dae94ef23610dfef83c1a5849
Apex can use codenames like "Q", "R" for its min_sdk_version property.
Also, cc_library can use codenames for its stubs.versions.
Bug: 152655956
Test: vendor/google/build/build_mainline_modules.sh
Change-Id: I077ad7b2ac5d90b4c8708921e43846206f05ba70
We actually want some targets (in particular, the MediaProvider apk)
to link against framework-mediaprovider.
The gist of what the neverallow rule achieves (disallow general use)
is achieved equally well with visibility rules, which
framework-mediaprovider already has.
Bug: 152891096
Test: m nothing
Change-Id: Ic4b0a571985b9ad1dfdd56d45035f224a622700e
Use droidstubs for public and system stubs to provide a list of @removed APIs. As these APIs are not present in the stubs, they are not whitelisted / greylised automatically. Keep them on greylist manually.
Bug: 143864733
Test: diff out/soong/hiddenapi/hiddenapi-flags.csv
Change-Id: I4c8e6899fadfdfd1da82f6f453cc92e71aa9b78c
Merged-In: I4c8e6899fadfdfd1da82f6f453cc92e71aa9b78c
Exempt-From-Owner-Approval: clean cherry-pick
(cherry picked from commit c7fb5c9964)
Reuse the generic Dist support instead of writing
custom mk rules for it. Use the new Tag support to
grab the classes.jar instead of the dexed jar.
Bug: 152618077
Test: m sdk dist
Change-Id: Id893d52cc24b66e6bd900b5062a59f3820000bcd
Make java_library support this mode of output, to allow
callers to dist the classes.jar file rather than the dexed
jar file.
Bug: 152618077
Test: followup CL
Change-Id: I5ba6949833a0fbb95376142aec5096ff5f084c00
We have known problems (docker; distros w/o user namespaces) which mean
that we won't be turning this into a fatal error anytime soon. Also
remove the bug report link, it's not useful to continue getting the same
reports.
Bug: 123210688
Test: treehugger
Change-Id: I271871d68150417ac938f074d3730cad4518e327
Merged-In: I271871d68150417ac938f074d3730cad4518e327
(cherry picked from commit 1871d88e4f)
For VNDK-Lite devices, which doesn't define BOARD_VNDK_VERSION, VNDK
APEX is built with only VNDK-Sp libraries with core variants.
Bug: 141908078
Bug: 152353068
Bug: 151635128
Test: TH
Change-Id: I0d08d32473368fd158818d4c2c72fc4cfad68ce6
This is a cherry-pick change.
Test: Built mainline module coverage data
Bug: 152117890
Change-Id: I47bf3e5d6e78c4518729bdb52616e248156d3cec
Merged-In: I47bf3e5d6e78c4518729bdb52616e248156d3cec
It's painful to debug without these.
Bug: 151160048
Test: do a build; find out -name '*.rsp'
Change-Id: I6fc25095442178c0bcbc1dea444ba51e9259c600
Merged-In: I6fc25095442178c0bcbc1dea444ba51e9259c600
A library may have both VNDK variant and stubs. getRefAbiDumpFile should
differentiate its vendor and core variants.
Bug: 152277104
Test: Add dumps to platform/ and vndk/ ; make libselinux.vendor libselinux
Change-Id: Iad038cf4cd3eccc3dfbef13fab67da044498ce77
Merged-In: Iad038cf4cd3eccc3dfbef13fab67da044498ce77
(cherry picked from commit 27d235f918)
Previously, when Q-targeting apexes are bundled-built, they are built
against the latest stubs.
It was because unwinder is linked dynamically in R and APIs are provided
by libc while Q apexes should run on Q where libc doesn't provide those
APIs. To make Q apexes run on Q device, libc++ should be linked with
static unwinder. But, because libc++ with static unwinder may cause problem
on HWASAN build, Q apexes were built against the latest stubs for bundled
build.
However, Q apexes should be built against Q stubs.
Now, only for HWASAN builds, Q apexes are built against the latest stubs
(and native modules are not linked with static unwinder).
Bug: 151912436
Test: TARGET_SANITIZE=hwaddress m
=> Q apexes(media, resolv, ..) are linked with the latest stubs
m
=> Q apexes are linked with Q stubs,
and Q apexes' libc++ is linked with static unwinder
Change-Id: If32f1b547e6d93e3955c7521eec8aef5851f908c
