| Commit message (Collapse) | Author | Age |
... | |
|
|
|
|
|
| |
Also make AspectKey#toString a little more informative.
PiperOrigin-RevId: 185600069
|
|
|
|
| |
PiperOrigin-RevId: 185595397
|
|
|
|
|
| |
RELNOTES:None
PiperOrigin-RevId: 185590627
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 185580415
|
|
|
|
|
|
| |
are referenced by TopLevelTargets when we discard the analysis cache.
PiperOrigin-RevId: 185574670
|
|
|
|
|
|
|
|
| |
parameter isn't present. Allows us to handle cases where the class type encodes the parameter value. This also gives a compile-time check that field is present before blindly using it in codec.
Lets us get rid of a non-AutoCodec class.
PiperOrigin-RevId: 185573686
|
|
|
|
|
|
|
| |
expr - the expression to be evaluated
word - the configuration (represented by the strings 'host', 'target', or 'null') to try to find the result(s) of 'expr' in. If some but not all results of expr can be found in the specified config, then the subset that can be is returned. If no results of expr can be found in the specified config, then an error is thrown.
PiperOrigin-RevId: 185572590
|
|
|
|
|
|
|
| |
Added a little javadoc and tests.
RELNOTES: None
PiperOrigin-RevId: 185569985
|
|
|
|
|
|
|
|
|
|
|
|
| |
I neglected to unify the message text before submit. This uses the Go version
for both go and Java.
R=laurentlb,mkanat
CC=blaze-firehose+cl
RELNOTES:
unifly lint glob(["**/*.java"]) message
PiperOrigin-RevId: 185569422
|
|
|
|
|
|
|
|
| |
Remove WalkableGraph#isUpToDate and BuildDriver#alreadyEvaluated and delegate the
work to implementation.
RELNOTES: None
PiperOrigin-RevId: 185562370
|
|
|
|
|
|
|
|
|
|
|
| |
The values (if present) are written into the manifest with this format:
Target-Label: <label>
Injecting-Rule-Kind: <kind>
In the future, JavaBuilder will make sure of this instead of command line arguments to find owners for jars for its add_dep commands.
PiperOrigin-RevId: 185557317
|
|
|
|
| |
PiperOrigin-RevId: 185556666
|
|
|
|
|
| |
RELNOTES:None
PiperOrigin-RevId: 185553306
|
|
|
|
|
|
| |
(Des|S)erializationContext.
PiperOrigin-RevId: 185547740
|
|
|
|
|
|
|
|
|
|
|
| |
ControlFlowChecker has a precondition check that assumes nested functions do not occur. While this assumption is reasonable for a valid Skylark syntax, linter can actually be invoked on malformed files and users would get a stack trace instead of a human-readable linter error.
Alternative and possibly a better strategy would to not run CFChecker in case parse errors are detected.
fixes #4511
Closes #4512.
PiperOrigin-RevId: 185538897
|
|
|
|
|
|
|
| |
try-with-resources code segments
RELNOTES:None
PiperOrigin-RevId: 185536875
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 185535264
|
|
|
|
| |
PiperOrigin-RevId: 185528799
|
|
|
|
|
|
|
| |
CppConfiguration
RELNOTES: None.
PiperOrigin-RevId: 185527875
|
|
|
|
|
|
|
|
|
|
| |
"blaze run --direct_run" so that the called binary knows about the working directory the client was called from.
Its cwd is its runfiles directory and if not for the fact that we have to convey *two* directories to it, I'd have considered changing that. As it is, however, we can't convey two directories with the cwd of the binary so we have to use environment variables.
RELNOTES[NEW]: "blaze run --direct_run" now exports the BUILD_{WORKSPACE,WORKING}_DIRECTORY variables to tell the binary about the cwd of the client and the workspace root.
PiperOrigin-RevId: 185515884
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add new flag called `--local_tmp_root`, which (if
specified) tells Bazel what temp directory should
locally executed actions use.
Fixes https://github.com/bazelbuild/bazel/issues/4621
Related to https://github.com/bazelbuild/bazel/issues/3215
RELNOTES[NEW]: The new "--local_tmp_root=<path>" flag allows specifying the temp directory for locally executed actions.
Change-Id: Ice69a5e63d0bf4d3b5c9ef4dbdd1ed1c5025f85e
PiperOrigin-RevId: 185509555
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
https://github.com/bazelbuild/bazel/issues/4625
What I thought was a short fix is turning into a long hunt, so I better roll this back to get the build green again.
I'm not yet 100% certain what the interactions are, but there's a chance that it's back to the drawing board.
*** Original change description ***
Fixing test-setup.sh occasionally missing stdout/stderr, on systems where "tail --pid" is supported.
The solutions aren't mine, the new test was taken from Ola's unknown commit and the way to avoid race condition courtesy of sethkoehler@
Mitigates #4608 for compatible Linux systems.
TESTED=manual scripts and new test case.
RELNOTES: None
PiperOrigin-RevId: 185482604
|
|
|
|
|
|
| |
Fixes #4614
PiperOrigin-RevId: 185474153
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 185455486
|
|
|
|
|
|
| |
Previously, RuleContext was referenced via the objc implementation of CppSemantics. Objects of that class are no longer held by CppCompileAction post-analysis.
PiperOrigin-RevId: 185446320
|
|
|
|
| |
PiperOrigin-RevId: 185441432
|
|
|
|
|
|
|
| |
compile time.
RELNOTES: None.
PiperOrigin-RevId: 185432867
|
|
|
|
| |
PiperOrigin-RevId: 185424287
|
|
|
|
|
|
|
|
| |
2) Also print the dependency issues to stderr.
3) Add golden tests to test the above two.
RELNOTES:n/a.
PiperOrigin-RevId: 185423338
|
|
|
|
| |
PiperOrigin-RevId: 185419668
|
|
|
|
|
|
|
|
|
| |
For now we will only block Java recursive globs. Any other languages or
extensions can be banned relatively easily.
RELNOTES:
Add lint check for discouraging glob(["**/*.java"])
PiperOrigin-RevId: 185417760
|
|
|
|
|
|
|
|
| |
to ConfiguredTarget.GetTarget(). Also remove equivalence requirements for
the ConfiguredTarget's target and the stored Target since there will soon no
longer be a Target in ConfiguredTarget.
PiperOrigin-RevId: 185417468
|
|
|
|
|
|
|
|
| |
This enables writing tests for android_instrumentation_test that mock
android_binary using a skylark rule that returns an AndroidInstrumentationInfo.
RELNOTES: None
PiperOrigin-RevId: 185417182
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 185412809
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ExperimentalEventHandler is in charge of composing the pretty progress bar
that Bazel displays. This progress bar is a multi-line message with control
characters printed on the terminal.
The progress bar was composed by issuing many individual writes to an
AnsiTerminal. Because the AnsiTerminal in this case was backed by an error
stream (which are unbuffered), each of these writes resulted in a gRPC to
the Bazel client to write the message to the console. gRPC calls are much
more expensive than calls to a file descriptor, and, in general, even small
writes to a file descriptor should be avoided when preparing long messages.
To fix this, fully buffer the output messages sent to the AnsiTerminal until
explicitly flushed. ExperimentalEventHandler was already doing the right
thing regarding flushes but did not account for the fact that each write
would be (unintentionally) sent directly to the terminal.
The flicker was significant: on a pathological case (building sandboxfs with
Bazel on my MacBook Pro 13" on macOS), this change shaves about 5 seconds of
build time on the previous 45 second-long build. I think this only happened
with "bazel run" and "bazel test" invocations and not "bazel build", but
I haven't really confirmed this.
RELNOTES: None.
PiperOrigin-RevId: 185405892
|
|
|
|
|
|
|
|
| |
We already intern the labels themselves. Benchmarks do not show any further
gain by interning the label names.
RELNOTES: None
PiperOrigin-RevId: 185394812
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
"tail --pid" is supported.
The solutions aren't mine, the new test was taken from Ola's unknown commit and the way to avoid race condition courtesy of sethkoehler@
Mitigates #4608 for compatible Linux systems.
TESTED=manual scripts and new test case.
RELNOTES: None
PiperOrigin-RevId: 185374273
|
|
|
|
|
|
|
|
| |
objc_library that it depends on.
See https://github.com/bazelbuild/bazel/issues/3352
PiperOrigin-RevId: 185371993
|
|
|
|
| |
PiperOrigin-RevId: 185370712
|
|
|
|
| |
PiperOrigin-RevId: 185369902
|
|
|
|
|
|
|
| |
The file can be generated during execution by a different rule.
RELNOTES:none
PiperOrigin-RevId: 185361140
|
|
|
|
|
|
|
|
| |
I don't think it's worth repeating things here. Let's point to the main
documentation.
RELNOTES: None.
PiperOrigin-RevId: 185356504
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In MSVC, `environ` is a macro (from `stdlib.h`):
```cpp
extern char*** __p__environ(void);
#define _environ (*__p__environ())
#define environ _environ
```
So `extern char **environ;` will be expanded as `extern char **(*__p_environ());` which is invalid. This causes compile warning on MSVC.
Closes #4487.
PiperOrigin-RevId: 185354631
|
|
|
|
| |
PiperOrigin-RevId: 185354353
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 185353994
|
|
|
|
|
|
|
|
|
|
|
| |
...even where they should be clear from the evaluation order.
Since reporting sometimes happens in different threads, there
might be races on the event bus. Explicit order constraints
allow the BuildEventStreamer to reorder those events correctly
in the case of a lost race.
Change-Id: Ib5211341c2016527e9268e8fdbd1677d4255b23c
PiperOrigin-RevId: 185345738
|
|
|
|
|
|
|
|
|
|
| |
Context implementations are currently empty, just doing the plumbing in this
change. Once this is in we can start passing along the ObjectCodecRegistry, which
will allow runtime codec resolution for classes not known at compile time.
We'll also inevitably add some memoization helpers, allowing us to optimize the
serialization process further.
PiperOrigin-RevId: 185305674
|
|
|
|
|
|
|
|
|
| |
android_local_test generates and R.class file and so this is necessary for projects that don't nest their BUILD files under a java/ or javatests/ root.
Fixes #4618
RELNOTES: None
PiperOrigin-RevId: 185281836
|
|
|
|
|
|
|
|
| |
- make Objects.requireNonNull and Long.compare rewrites compatible with --core_library
- apply those and try-with-resources rewrites to generated companion classes
RELNOTES: None.
PiperOrigin-RevId: 185262256
|