This reverts commit 782b752889.
bug: 308932171
tested: locally generated and applied ota on nonab device
Change-Id: I67b26be01d6240b2cb332e13bc7c7959de128755
Rather than use an unsupported flag setting that the user likely
doesn't even realize is being used, we immediately stop the build.
This error message is more verbose, mentioning 'lunch', because
it's anticipated a lot more users will hit this issue when first
switching to trunk stable, and more details will hopefully help
them out.
We have some complication in that some internal commands set
TARGET_RELEASE to an empty string. We put in logic to allow
that path.
Since $(error) immediately stops the build, we also get rid of
some 'else' logic and indentation, to hopefully offset some of
the complication we've added.
Bug: 307946156
Test: 'lunch' (still) works; A build attempt without `TARGET_RELEASE` set (now) fails
Change-Id: I2f667668c6688e501a3536981b7bae5fdbf962a0
These files don't have anything to do with bazel, they just use starlark as a configuration
language. Bazel recently introduced the scl file extension to use for this format, which doesn't
have any bazel-specific symbols. Use that extension for our pure starlark files as well.
Bug: 309686282
Test: Presubmits
Change-Id: I7b08f342e7fb94405a52af0918ae6a7d542f3282
- Introduce a new directory for some specific proguard rules.
kotlin.flags:
- Remove DebugMetadata
- Don't warn about missing non-runtime annotations
Test: build AOSP & test on internal
Bug: 309023911
Change-Id: Iee740b61a2afeba3482ff4e1f8213bd4cf46174a
.scl files are starlark configuraiont language files, which is starlark
without any of the bazel buildin symbols.
This is to match bazel's behavior, and it doesn't make sense for
scl files to load bzl files.
We should change all the release config bzl files to scl, and then
also make it so that when using executionModeScl you can only load
scl files.
Test: go test
Change-Id: I196bcf3a4548118791ed1d47c2d37e46a1ef86c4
Allow a build flag definition to indicate that its value should be the
concatentation of assignements, rather than the final assigned value. In
this case, the "default" value from the flag definition is always
present as the start of the list.
The initial use case for this is RELEASE_ACONFIG_VALUE_SETS, where we
need apply multiple definition files that should be processed to arrive
at the final value.
Bug: b/302593603, b/304814040, b/309477343
Test: manual
Change-Id: I58eb71f2ee6d8f08f11a432993f23157831ec93c
1. release config maps now specify where the flag definitions are found.
2. PRODUCT_RELEASE_CONFIG_MAPS specifies additional release config map
files to use.
This allows product config to specify build flags, which can then be
specified by users of that product.
Bug: b/302593603, b/309477343
Test: manual
Change-Id: Ic1f0512ec4b06ac94dd3f29eadd6a03ba8ebf6d2
Previously, the 16K/4K boot options OTAs are full OTAs, resulting in
file size of ~20MB each, and ~40M for both OTAs. To reduce the space
usage, use incremental OTAs instead.
Test: th
Bug: 302759296
Bug: 293313353
Change-Id: I61cc84c6c13f151dd6bc5ff37dd31daa5fb31abd
This is a new mechanism for asserting properties about your product
config. See the documentation in product_validation_checks.mk for
more information.
Test: Manually
Change-Id: I698dea899441f3773f839ea2ba1a2a6cfe59b57b
they were moved to device/generic/goldfish
Bug: 295259752
Test: presubmit
Change-Id: I996811e8703b43313fca3ac819e6ed333d37866b
Signed-off-by: Roman Kiryanov <rkir@google.com>
* changes:
Remove code related to unused LOCAL_* variables
Delete unused variables from clear_vars.mk
Remove obsolete ide.mk and related code
Removed unused license code
Remove obsolete uses of LOCAL_MODULE_TAGS
`ot-cli-ftd` is a CLI tool simulating a node in Thread network. See https://www.threadgroup.org/ about Thread networking technology.
It's currently in `/vendor/bin/` but we're moving it to `/system/bin/` because `ot-cli-ftd` will be used in various CI tests.
`ot-cli-ftd` is also useful for debugging purpose.
Bug: 309051262
Test: Verified on Cuttlefish.
Change-Id: I96bf2bd7575540dc86e969a80cd9bb5634c3a1ea
LOCAL_JETIFIER_ENABLED, LOCAL_NO_PIC, LOCAL_PREBUILT_STRIP_COMMENTS,
LOCAL_RMTYPEDEFS, and *.vts sources are never used. Remove them
and the code related to them.
Test: no change to out/build-aosp_cf_x86_64_phone.ninja
Change-Id: I2ca9e674602057cc163b8bc28b0c57a0b7cc4361
Kati analysis in AOSP spends around 6 seconds in clear_vars.mk.
Delete any variables in clear_vars.mk that are not referenced anywhere
else in build/make/core.
Test: no change to build-aosp_cf_x86_64_phone.ninja
Change-Id: I7e0db3c02d297de825acbfbd1a0f05724d1e846d
The code to generate an eclipse classpath is obsolete, remove it and
related code.
Test: no change to out/build-aosp_cf_x86_64_phone.ninja
Change-Id: I7e1b6268b98ecbb7be88db8945dd7b30acc695ba
Nothing ever reads ALL_DEPS.*.LICENSE, and its an ever-growing list
that is sorted every time, which is extremely expensive.
notice_target is never set after I4cddf9a381a1258bdc2b1b42be72c447df10d234,
remove all the related code.
Test: no change to out/build-aosp_cf_x86_64_phone.ninja
Change-Id: I0fa6a46e62ef8aa78873b43d3064b57b1c54de51
Only LOCAL_MODULE_TAGS := tests is used is meaningful in the current
implementation. "optional" and "samples" both exist in the tree, but
are meaningless. "gnu", "user", "eng" and "debug" are no longer used,
and are already forbidden by the unusual tags check. The info from the
"module" target is now available in module-info.json. Delete all the
irrelevant code.
Test: no change to out/build-aosp_cf_x86_64_phone.ninja
Change-Id: I04e8178a362e382a1a4bd997c1b4c3a480db7714
