| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Causes crashes in fake_cc_binary in combination with --experimental_nodotd_scanning_with_modules.
*** Original change description ***
Stop generating .d files when they won't be looked at later.
While at it remove CppCompilationActionBuilder.setAllowUsingHeaderModules,
which isn't used anymore and would make the logic here (even) more complicated.
RELNOTES: None.
PiperOrigin-RevId: 207932548
|
|
|
|
|
|
|
|
|
|
|
|
| |
This class just contains methods used elsewhere; move them to appropriate
places (generally AndroidManifest or StampedAndroidManifest).
Also, somewhat reduce code duplication by having more stuff use the existing
AndroidManifest.from() method, which automatically handles unspecified
manifests and packages.
RELNOTES: none
PiperOrigin-RevId: 207913410
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
1) Break up dense lines with clearer pretty-printing.
2) When a violation happens because of a select(), mention
both the target with the select (as before) *and*
the dep that the select() chose.
3) Integrate this messaging into --target_environment violations,
which currently provide no info about the root cause.
Examples:
-------------------------------------
select() + compatible_with violation:
-------------------------------------
Before:
ERROR: /workspace/testapp/BUILD:41:1: in cc_binary rule //testapp:top: the current command-line flags disqualify all supported environments because of incompatible select() paths:
environment: //constraints:p removed by: //testapp:midlib (/workspace/testapp/BUILD:28:1)
After:
ERROR: /workspace/testapp/BUILD:41:1: in cc_binary rule //testapp:top: the current command line flags disqualify all supported environments because of incompatible select() paths:
environment: //constraints:p
removed by: //testapp:midlib (/workspace/testapp/BUILD:28:1)
which has a select() that chooses dep: //testapp:glib
which lacks: //constraints:p.
-------------------------------------
select() + --target_environment=//constraints:p violation:
-------------------------------------
Before:
ERROR: This is a restricted-environment build.
- //testapp:top does not support required environment //constraints:p
After:
ERROR: This is a restricted-environment build.
//testapp:top does not support:
environment: //constraints:p
removed by: //testapp:midlib (/workspace/testapp/BUILD:28:1)
which has a select() that chooses dep: //testapp:g
which lacks: //constraints:p
Fixes: #5795
PiperOrigin-RevId: 207910308
|
|
|
|
|
|
| |
Bazel's ASM was updated in 3a711882dcbb3af8709844bde501ac6fca44ea7d.
PiperOrigin-RevId: 207909203
|
|
|
|
|
|
| |
tests. This could change in the future but has confused several people recently so documenting the state of the world as is for now.
PiperOrigin-RevId: 207908818
|
|
|
|
|
|
|
|
| |
This was also just shared between ParsedAndroidResources and the now defunct
ResourceContainer
RELNOTES: none
PiperOrigin-RevId: 207907140
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 207905848
|
|
|
|
|
|
|
|
|
|
|
| |
expressions.
This is probably only a theoretical problem, since a blocking struct field is probably a very bad idea.
Closes #5132.
Change-Id: Ie84a78ab4d9ce215f2806ac49bf8911de6959930
PiperOrigin-RevId: 207902766
|
|
|
|
|
|
|
|
| |
Now that ValidatedAndroidResources is the only implementation, we can just use
that instead.
RELNOTES: none
PiperOrigin-RevId: 207900844
|
|
|
|
|
|
|
|
|
| |
instead, read the heads from the configured java runtime, so the values
track with changes to --javabase.
See #5594
PiperOrigin-RevId: 207894086
|
|
|
|
| |
PiperOrigin-RevId: 207891979
|
|
|
|
|
|
|
|
|
| |
ValidatedAndroidResources is now the only implementation of
ValidatedAndroidData, so we can also clean up some code. (The actual interface
will be cleaned up in the next few changes.)
RELNOTES: none
PiperOrigin-RevId: 207891778
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes https://github.com/bazelbuild/bazel/issues/5751
RELNOTES[INC]: Windows: when BAZEL_SH envvar is not defined and Bazel searches for a suitable bash.exe, Bazel will no longer look for Git Bash and no longer recommend installing it as a Bash implementation. See issue #5751.
Change-Id: I7350b1dd5c0a3777525956da6d620174fc6935ee
Closes #5752.
Change-Id: I7350b1dd5c0a3777525956da6d620174fc6935ee
PiperOrigin-RevId: 207891008
|
|
|
|
|
|
|
|
|
|
|
| |
instead of indirecting through javac's bootclasspath handling and attempting
to pin to a particular source version. This is a stop-gap until we can just
use javac's --release flag.
Using the output of DumpPlatformClasspath as the bootclasspath for the default
java_toolchain side-steps issues with @local_jdk (see #5744, #5594).
PiperOrigin-RevId: 207890272
|
|
|
|
|
|
|
|
|
|
| |
//tools/jdk:current_java_runtime
instead of referencing @local_jdk directly.
See bazelbuild/bazel#5594
PiperOrigin-RevId: 207889762
|
|
|
|
|
|
|
| |
the DTC construction of the universe argument to 'rdeps'.
RELNOTES: None
PiperOrigin-RevId: 207884395
|
|
|
|
|
|
|
| |
use Semaphore#availablePermits. I have no idea why I didn't do this initially.
RELNOTES: None
PiperOrigin-RevId: 207883650
|
|
|
|
| |
PiperOrigin-RevId: 207882126
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Remove all state from the ApplicationManifest class. Actual methods in this
class will be migrated to somewhere more appropriate in upcoming changes.
We now usually use the manifest passed around in the ResourceApk object
instead. (The exception is mobile-install, which still wants to use the raw
manifest.) Note that this manifest is the processed manifest output by resource
processing - using it will somewhat decrease parallelisim, but is also more
correct. (We'll get the parallelisim gains back once we now longer use the
legacy manifest merger and don't need to process the manifest in resource
processing.)
RELNOTES: none
PiperOrigin-RevId: 207881830
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If a Bazel server is idle for 10 seconds, it unconditionally triggers a full-scale Java GC via System.gc(). This behavior doesn't have clear benefits and causes Bazel to steal resources from whatever the user does after invoking Bazel. This CL adds a startup option, --idle_server_tasks, to toggle the idle GC behavior.
Also, add some logging for when idle GC is enabled, so it's easier to evaluate its effects. Example of logging:
```
180718 17:43:04.609:I 247 [com.google.devtools.build.lib.server.IdleServerTasks.lambda$idle$0] [Idle GC] used: 157MB -> 15MB, committed: 421MB -> 422MB
```
Fixes https://github.com/bazelbuild/bazel/issues/5589.
Closes #5628.
PiperOrigin-RevId: 207869996
|
|
|
|
|
|
|
|
|
| |
Cloning branches as added after the fact to the skylark git rules,
once recording the actual commit was implemented. Still support shallow
cloning also when following branches.
Change-Id: Ib7a84de4ada69152e9e7c27e9d84cff39cfd3582
PiperOrigin-RevId: 207866303
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Enable the tests in //src/test/shell/integration:*
to run on CI, and fix them all.
See https://github.com/bazelbuild/bazel/issues/4292
Change-Id: I6a811d8045d85237a7df4024f893f3ca1b25177a
Closes #5789.
Change-Id: I6a811d8045d85237a7df4024f893f3ca1b25177a
PiperOrigin-RevId: 207863832
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This cl exposes getters for flags passed by --copt, --cxxopt, --conlyopt, and
--linkopt Bazel options.
Fixes #5602.
While at it, I deprecated existing methods that are either migrated to the C++
toolchain, or to cc_common.
RELNOTES: None.
PiperOrigin-RevId: 207854692
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 207850273
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 207845945
|
|
|
|
| |
Fixes #5797
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 207801155
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 207778688
|
|
|
|
| |
PiperOrigin-RevId: 207773685
|
|
|
|
|
|
|
| |
Built at:
https://github.com/google/turbine/commit/598b467aa1cf3fac39878ab54a56af11629e4079
Fixes #5782
|
|
|
|
|
|
|
|
| |
Also, migrate Robolectric action that used it to use AndroidDataConverter
instead.
RELNOTES: none
PiperOrigin-RevId: 207761747
|
|
|
|
|
|
|
|
| |
This code should all be unused now. Some code it calls into will be removed in
the next changes.
RELNOTES: none
PiperOrigin-RevId: 207753966
|
|
|
|
|
|
| |
RELNOTES:
Replace 0/1 with False/True for testonly attribute
PiperOrigin-RevId: 207747213
|
|
|
|
|
|
|
|
|
|
|
| |
and genquery non-determinism.
This approach relies on there never being two attributes with the same name.
I added the inOrder() to the test without my change commented and observed flaky failures. I then uncommented my change and saw consistent passes.
RELNOTES: Sort attribute lists in proto-form query output to fix non-deterministic genquery output.
PiperOrigin-RevId: 207743773
|
|
|
|
|
|
|
|
| |
Unlike previously, this is not magically selected when running on JDK 8.
Use --java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8
--host_java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8 to configure
the JDK 8 host_javabase-compatible toolchain.
PiperOrigin-RevId: 207738992
|
|
|
|
| |
PiperOrigin-RevId: 207734653
|
|
|
|
|
|
|
|
|
|
| |
This flag is turned on everywhere. Remove it.
There's a lot of dead code hidden behing this flag; will remove it in a series
of upcoming changes.
RELNOTES: none
PiperOrigin-RevId: 207732126
|
|
|
|
| |
PiperOrigin-RevId: 207730551
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 207724147
|
|
|
|
|
|
|
| |
bound check as it is intermittently failing under very heavy load.
RELNOTES: None.
PiperOrigin-RevId: 207716645
|
|
|
|
|
|
|
|
|
|
| |
set of all declared headers, iterate over them and remove them from a (often
much smaller) set.
This should improve runtime as well as amount of garbage produced.
RELNOTES: None.
PiperOrigin-RevId: 207710867
|
|
|
|
|
|
|
|
| |
https://github.com/bazelbuild/bazel/issues/5633
Closes #5647.
PiperOrigin-RevId: 207706657
|
|
|
|
|
|
|
|
| |
Use `git -C` instead of `cd` in git skylark rules to be more concise. I think it's probably better to use the built-in changedir functionality of git rather than `cd`ing around manually. Happy to be overruled however if it makes readability better.
Closes #5668.
PiperOrigin-RevId: 207703326
|
|
|
|
|
|
|
|
|
|
| |
Side-note: `options_test`, `zip_headers_test` and `zlib_interface_test` now pass on Windows.
/cc @laszlocsomor
Closes #5781.
PiperOrigin-RevId: 207699280
|
|
|
|
|
|
| |
*** Reason for rollback ***
PiperOrigin-RevId: 207697096
|
|
|
|
|
|
|
| |
Paths to tools in CROSSTOOL are either absolute or relative to the CROSSTOOL location (which is the same as cc_toolchain location). As in the future CROSSTOOL will be gone, and the new skylark rule that will replace CROSSTOOL will not have to be in the same location as cc_toolchain, we need to pass information to FeatureConfiguration about the location of the cc_toolchain in use, so we can calculate the workspace relative paths to the tools.
RELNOTES: None.
PiperOrigin-RevId: 207695703
|
|
|
|
|
|
|
|
| |
/cc @laszlocsomor
Closes #5779.
PiperOrigin-RevId: 207693540
|
|
|
|
|
|
|
|
|
|
| |
git_repository could use HTTP_PROXY to clone repository
Fixes #3893.
Closes #5674.
PiperOrigin-RevId: 207689348
|
|
|
|
|
|
|
| |
All users have been migrated, so it's not used anymore.
RELNOTES: None.
PiperOrigin-RevId: 207688604
|
|
|
|
|
|
| |
now that JDK 8 host_javabases are no longer supported.
PiperOrigin-RevId: 207687089
|