| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
| |
crosstool top.
Change-Id: I531034b0c991d18b05818db4b40cbd739535b565
PiperOrigin-RevId: 187617580
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fix the failing test in
//src/test/java/com/google/devtools/build/lib:bazel-rules-tests
on Windows.
The problem was that the test setup read a
resource file and assumed line endings to be "\n",
but on Windows this resource file was generated
with "\r\n" endings.
The test setup code then put those strings into a
set.
Finally, the test tried to match a string in this
set, but because the entries had an invisible "\r"
at the end, the matching failed.
This commit trims the entries before putting them
in the set.
Fixes https://github.com/bazelbuild/bazel/issues/4752
Closes #4754.
PiperOrigin-RevId: 187608803
|
|
|
|
|
| |
RELNOTES: Use bazel dump --action_graph=/path/to/action.proto --action_graph:targets://foo:bar,//foo:foo to filter for certain targets in the action graph dump.
PiperOrigin-RevId: 187608321
|
|
|
|
|
|
|
|
| |
The previous comment referred to the option `/EHa`, but `/EHsc` is the option to only catch C++ standards (which seems to be what we want, according to this doc: https://msdn.microsoft.com/en-us/library/1deeycx5.aspx)
Closes #4739.
PiperOrigin-RevId: 187593737
|
|
|
|
|
| |
RELNOTES:none
PiperOrigin-RevId: 187591225
|
|
|
|
|
|
|
| |
Fixed https://github.com/bazelbuild/bazel/issues/4723
RELNOTES:
PiperOrigin-RevId: 187589378
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187571707
|
|
|
|
|
|
|
|
|
| |
Instead of --direct_dependency, --indirect_dependency args we now emit --direct_dependencies. We no longer need to emit any jar owner information since that is baked into the jar by JavaBuilder.
This CL also contains the deletion of CustomMultiArgv and the injecting_rule_kind aspect parameter, as the deleted code was the last remaining usage.
RELNOTES: None
PiperOrigin-RevId: 187558628
|
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187551970
|
|
|
|
|
|
|
|
|
| |
objc provider skylark fields.
This flag is not fully implemented -- there should also be a way to disable the provider getters using this flag, but, since @SkylarkCallable does not allow for evaluation of SkylarkSemantics, this is not yet possible.
RELNOTES: None.
PiperOrigin-RevId: 187542004
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 187539005
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187531006
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 187530224
|
|
|
|
|
|
|
| |
(It's good to avoid passing around more information than one needs, and FuncallExpression is pretty heavy when only a Location is necessary)
RELNOTES: None.
PiperOrigin-RevId: 187529568
|
|
|
|
|
|
|
| |
(https://github.com/google/copybara/commit/77a0e68eb0b044449b412e939ad25d500d2c12d3).
RELNOTES: None
PiperOrigin-RevId: 187525928
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
With this cl toolchain author can specify different flags for linking shared
library produced by cc_library and a shared library produced by cc_binary.
This is what is needed to remove linking_mode_flags - MOSTLY_STATIC_LIBRARIES
from the crosstool. What this linking mode was used for was to separate when we
link transitive shared library from cc_binary and when we link this
little-and-not-really-useful-outside-of-bazel nodeps shared library in cc_library.
RELNOTES: CcToolchain: Introduced action_config for "c++-link-transitive-dynamic-library"
PiperOrigin-RevId: 187523334
|
|
|
|
|
|
|
| |
fix logic for implementing emulated interfaces.
RELNOTES: None.
PiperOrigin-RevId: 187520298
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previous behaviour:
* The CopyAndRewrite function would raise a write permission denied error on the output file, which had permissions 0555, when the input SPEC file was auto-generated.
Current behaviour:
* Do not copy the input SPEC file (auto-generated or otherwise) to the destination in the temporary working directory since the CopyAndRewrite function will generate the output file regardless. This allows the package to support static or auto-generated SPEC files.
Closes #4716.
PiperOrigin-RevId: 187511568
|
|
|
|
|
|
|
| |
a nice consequence, this lets us reduce GC churn since we no longer need to create a frame instance for the lexical frame at a callsite of either a function when the environment is frozen or a builtin function (since builtins cannot modify bindings in their lexical frame).
RELNOTES: None
PiperOrigin-RevId: 187495787
|
|
|
|
|
|
|
|
| |
The changelog is not used by make_rpm.py, and I don't have a need for it since I auto-generate the SPEC file, which includes a "%changelog" section, prior to invoking the pkg_rpm rule.
Closes #4717.
PiperOrigin-RevId: 187494400
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187489449
|
|
|
|
|
|
|
| |
The links were to different sections in the page, but were not prefixed with hashtags.
RELNOTES: Fix broken links in query docs.
PiperOrigin-RevId: 187486286
|
|
|
|
|
|
| |
Closes #4712.
PiperOrigin-RevId: 187486082
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The output files are created without a repository, but the expected
filenames have them
This resolves issues when having a proto_library from an external build
file.
This seems to be a regression, so maybe should go into the 0.8.0 branch?
Note: Work at Square and we have a signed CLA with google
Note, without this fix we get errors like
```
ERROR: /private/var/tmp/_bazel_lewis/4a25cfc2b9b758043413ac58525ef6b4/external/AllProtos/BUILD.bazel:27:1: output 'external/AllProtos/squareup/objc/objc.j2objc.pb.m' was not created
ERROR: /private/var/tmp/_bazel_lewis/4a25cfc2b9b758043413ac58525ef6b4/external/AllProtos/BUILD.bazel:27:1: output 'external/AllProtos/squareup/objc/objc.j2objc.pb.h' was not created
```
Closes #4058.
PiperOrigin-RevId: 187480864
|
|
|
|
|
|
| |
RELNOTES:
Removed flag `--incompatible_load_argument_is_label`.
PiperOrigin-RevId: 187479614
|
|
|
|
|
|
|
| |
For a filter on option values (either by whitelist, allow_values, or blacklist, disallow_values), one of the options for what to do when encountering a disallowed value is to replace it with the default. This default must be itself an allowed value for this to make sense, so this is checked. This check, however, shouldn't apply to flags that are null by default, since these flags' default value is not parsed by the converter, so there is no guarantee that there exists an accepted user-input value that would also set the value to NULL. In these cases, we assume that "unset" is a distinct value that is always allowed.
RELNOTES: None.
PiperOrigin-RevId: 187475696
|
|
|
|
|
|
|
|
|
|
|
| |
The test_errmsg test tested the behaviour when workspace_status_command
was set to a file that did not exist, but other tests created a file
with that name. If test_errmsg did not run first, the
workspace_status_command would be valid and the test would fail.
Closes #4704.
PiperOrigin-RevId: 187470549
|
|
|
|
| |
PiperOrigin-RevId: 187467154
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The skylark implementation of git_repository is a drop-in replacement
for the native workspace rule. So deprecate the latter to avoid having
two implementations of the same. Also, the native rule pulls in quite
some dependencies.
RELNOTES: The native git_repository rule is deprecated. Use the
Skylark version available via load("@bazel_tools//tools/build_defs/repo:git.bzl", "git_repository")
instead.
Change-Id: Id75d8e7b7377acd1092c4ba993f488be86c03fa0
PiperOrigin-RevId: 187466931
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit c8be465869fbcfaa00b75d241c67279324976e0b added the headless JDK
to `debian/control`, but this isn't propagating to published packages
because those use the deps list in `debian/BUILD`.
This commit adds the headless JDK to `debian/BUILD`, with comments so
future changes won't hit the same problem.
Closes #4686.
PiperOrigin-RevId: 187461399
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187453379
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Google C++ Style Guide disallows static variable with non-trivial
destructor. Use constexpr for static variable whenever possible.
std::array would make the code cleaner, but MSVC's implementation
of constexpr std::array is buggy.
Side-benefit: linear search is faster when array is small (<20).
Change-Id: Ic8244dcb868e27d02ceb5298ccec482e7d4254b7
PiperOrigin-RevId: 187451495
|
|
|
|
|
|
|
|
|
|
| |
By simply sharing the utility function. In this way, we get
feature parity between git_repository and http_archive.
Fixes #4676.
Change-Id: I99b300e42b2f267d8d04fd965f09c24f3ae54f10
PiperOrigin-RevId: 187450644
|
|
|
|
|
|
|
| |
Fixed https://github.com/bazelbuild/bazel/issues/4722
RELNOTES: None
PiperOrigin-RevId: 187443457
|
|
|
|
|
|
|
| |
installBase is also used to access the embedded binaries.
RELNOTES: None
PiperOrigin-RevId: 187422136
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The plugin tries to read the following attributes:
Target-Label: <label>
Injecting-Rule-Kind: <rule kind>
These are used instead of the jar owners passed to JavaBuilder from blaze. Any owners currently passed from blaze are now ignored.
This CL also half-reverts the command line syntax for JavaBuilder and Turbine. It is now:
--classpath <jar1> <jar2> ...
--direct_dependencies <jar1> ...
RELNOTES: None
PiperOrigin-RevId: 187419469
|
| |
|
|
|
|
| |
PiperOrigin-RevId: 187409187
|
|
|
|
|
|
|
|
| |
The current behavior is already correct, just adding a test to make sure we retry reads as we should.
TESTED=the unit test
RELNOTES: None
PiperOrigin-RevId: 187398578
|
|
|
|
| |
PiperOrigin-RevId: 187397314
|
|
|
|
|
|
| |
More strictly type the arguments for creating FileTypes.
PiperOrigin-RevId: 187391336
|
|
|
|
|
|
| |
duplicates of those attributes in query's proto output format.
PiperOrigin-RevId: 187387578
|
|
|
|
|
|
|
| |
Logs whenever BuildEvenServiceTransport starts a PublishBuildToolEventStream
streaming RPC.
PiperOrigin-RevId: 187374228
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 187370833
|
|
|
|
|
|
| |
for every type of action.
PiperOrigin-RevId: 187368369
|
|
|
|
|
|
|
|
| |
output to given attributes.
RELNOTES[NEW]: blaze query: use --proto:output_rule_attrs to filter for given attributes
PiperOrigin-RevId: 187364996
|
|
|
|
|
|
|
|
|
|
| |
These subclasses should be using @SkylarkCallable(structField = true) instead
This is a bit of a memory win, as there is now no need to store field information twice.
There are still a couple of stragglers that are more difficult, namely ToolchainInfo and DefaultInfo. Their APIs will likely need some more extensive revamping before proceeding.
RELNOTES: None.
PiperOrigin-RevId: 187364392
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 187356329
|
|
|
|
|
|
|
| |
initialization of the Predicates, and storing the set of strings the
predicates are based on.
PiperOrigin-RevId: 187351700
|
|
|
|
|
|
| |
Remove the HashCode marshaller.
PiperOrigin-RevId: 187350917
|