The path to the artifacts will in the form of
/system/framework/oat/<arch>/<encoded-jar-path>@classes.{odex,vdex,art},
where <encoded-jar-path> is the path to the jar file with "/" replaced
by "@". For example,
/system/framework/oat/x86_64/apex@com.android.art@javalib@service-art.jar@classes.odex
There will be a follow-up CL to update ART runtime to recognize
artifacts in that path.
Test: m com.android.art
Bug: 194150908
Change-Id: Ic89fd63c4b1cd565684cead83fc91dae3bc97a4c
dep.OutputFile().Path() can panic if the OutputFile
is invalid, so move more of the code into the check
for validity. This changes the go crash into a
build error.
Bug: 158624492
Test: Apply the patchset 1 of ag/11815659, lunch aosp_x86_64-eng, m
Change-Id: Ic8a728be1b570ac73be7b2b9461676f9fd99c266
`m product-graph` runs a "non-full" build, which skips
some steps to improve performance. One of the steps
it skips is making a .installable_files file. Soong
reads this file and removes any files that were removed
from the list in this file. It errored out if the file
didn't exist. Make soong less strict about the presence
of this file, because if it doesn't exist, there shouldn't
be any installed files to remove either.
Fixes: 168105598
Test: rm -rf out/, m product-graph
Change-Id: I366f7b09d87911f9660d4e08c2d2f097cc04800f
Technically, tzdata apex depends on
//system/sepolicy/apex:com.android.tzdata-file_contexts. Builds suceeded
regardless due to bp2build generating a filegroup for all files in a
directory.
This also requires an update to the check in apex/builder.go to address
the fact that outputs (including filegroups) from Bazel in mixed builds
will be in a Bazel out subdirectory. This change also corrects a
potential bug that would not support all subdirectories under
system/sepolicy as golang's path.Match does not treat ** as recursive.
Test: go test apex tests
Test: build/bazel/ci/mixed_droid.sh
Change-Id: I1076d60cf271e4cdd1ea990156ab0a39fcfa6535
... for native tests. This new location has the appropriate
settings for debuggerd to generate valid backtraces in case
of test binary crash.
Bug: 167308193
Bug: 199904562
Test: compile
Change-Id: Ic106011a79f0aec86896b305616cf5590ab2eb2d
In particular it's useful to print the module, since the panic is
delayed from the init call.
#codehealth
Test: m nothing
Change-Id: Ia91127be83d8a9ce08cf0c08bc3f13fce7be14af
Test: Ran the python version of the script and the python3 version
Test: and verified the output is the same. This is for all three
Test: outputs (html, text, xml).
Test: Builds without an error.
Change-Id: I344504a4351c14e00caeb109c5507a60fb6be05a
Add new module:
host_snapshot {
name: "host-snapshot"
deps: [
(list of host tools)
],
...
}
Package host tools using android.PackagingBase to capture host tools and transitive packaging data.
Add JSON meta data to snapshot that allows snapshot to be installed via development/vendor_snapshot/update.py
Add support to generate a fake host snapshot of all host modules that is used to detect required
modules via development/vendor_snapshot/update.py.
Bug: 192556798
Bug: 194799048
Bug: 192896149
Test: m HOST_FAKE_SNAPSHOT_ENABLE=true host-fake-snapshot dist -- check snapshot exists in dist
Change-Id: I849c4db801cd858408f6fe6a3ce69262a23a5be9
Previously these were expanded into copts, requiring making all includes
absolute and duplicating includes to account for potentially generated
files. We now can handle both of these properly on the Bazel side, so
let's clean up build files a bit.
Test: bp2build.sh
Change-Id: I6c6160738cd6c269408c6c7a37010654d84f3c9d
We turned this off in local tidy flags, but didn't disable it on global
tidy builds. Let's remove this for now, since the upkeep cost is large
on it, and it also is currently crashing in some cases.
Bug: http://b/199534745
Bug: http://b/194865868
Test: WITH_TIDY=1 m
Change-Id: Ife39ab4949c5baf825efff66326b276c6a2a59f4
Bug: http://b/199203503
Upstream LLVM change 6d2d3bd0a6 made -z,start-stop-gc the default. It
drops metadata sections like __llvm_prf_data unless they are marked
SHF_GNU_RETAIN. https://reviews.llvm.org/D97448 marks generated
sections, including __llvm_prf_data as SHF_GNU_RETAIN. However this
change is not in the Rust toolchain. Since we link Rust libs with new
lld, we should use nostart-stop-gc until the Rust toolchain updates past
D97448.
Test: m unicode-xid_device_test_src_lib and verify that they have the
__llvm_prf_data section.
Change-Id: I55eb3622dae9bd789fbacfe708600316102c365f
Rust builds were picking up the crt objects and system libraries like
libc.so from /usr/lib/x86_64-linux-gnu/Scrt1.o and
/lib/x86_64-linux-gnu/libc.so.6. Pass --sysroot to the linker to
point it to the glibc prebuilts.
Test: manual
Change-Id: I6540df8aef0e5c2258de77456d18a6052db627d4
* Tested with a clang-tidy wrapper that dumps TIDY_TIMEOUT.
Test: make WITH_TIDY=1 TIDY_TIMEOUT=42
Bug: 199451930
Change-Id: I86580225bad7487da786db2ba62c09bd99d70c67
The specification of exporting includes vs system includes has an impact
on inclusion sort order. Conflating the two caused some symbols to not
be resolved correctly.
Bug: 198403271
Test: build/bazel/ci/bp2build.sh
Test: USE_BAZEL_ANALYSIS=1 m libbacktrace_no_dex succeeds with libc++_*
modules removed from mixed build denylist (would fail otherwise)
Change-Id: I08aff253d8962dc678ed10214b1c171330e0fe19
Please do not roll back.
If you have recently created a new Soong module type, please make sure
it supports the `licenses` property.
If your build on an older branch fails due to this change, please
configure your build to set `ANDROID_REQUIRE_LICENSES=false` in the
environment.
If running from the command line, the following will work:
export ANDROID_REQUIRE_LICENSES=false; m -j ...
Previously defaulted to not require property unless overridden in env.
Bug: 151177513
Test: m all
Change-Id: Ib295658f978511d07197c295f04a6f25f7d83686
