aboutsummaryrefslogtreecommitdiffhomepage
Commit message (Collapse)AuthorAge
...
* Replace usage of foo.com with example.com in git_repository test dataGravatar Ed Baunton2018-04-09
| | | | | | | | Fixes https://github.com/bazelbuild/bazel/issues/4409 Closes #4522. PiperOrigin-RevId: 192112573
* Remove BuildConfiguration.Fragment#getReservedActionMnemonics() in favor of ↵Gravatar lberki2018-04-09
| | | | | | | | | encoding the same in ConfiguredRuleClassProvider. This is a step towards dumbing down BuildConfiguration.Fragment and the ConfigurationFactoryLoader, which is in needed so that we can rewrite C++/Java/Python rules in Skylark without having to introduce the concept of "configuration loader" in Skylark, too. RELNOTES: None. PiperOrigin-RevId: 192104912
* Made 'file_posix.cc' POSIX compatibleGravatar Jasper Siepkes2018-04-09
| | | | | | | | | | | | | | | | The `d_type` field is not part of the POSIX specification. Added a compile time check to see if we can use it. When not present fallback to a (slightly more expensive) call to 'stat'. The reason why I need this is because I'm trying to port Bazel to a POSIX compliant platform. Even though my porting effort might fail miserably ;-) I think having a POSIX implementation which is POSIX compliant would be of benefit to the Bazel project. Besides the POSIX spec I've based the implementation on information I found here: * https://stackoverflow.com/questions/2197918/cross-platform-way-of-testing-whether-a-file-is-a-directory * https://stackoverflow.com/questions/23958040/checking-if-a-dir-entry-returned-by-readdir-is-a-directory-link-or-file#answer-29094555 * https://stackoverflow.com/questions/39429803/how-to-list-first-level-directories-only-in-c/39430337#39430337 Closes #4967. PiperOrigin-RevId: 192102522
* Made 'build-runfiles.cc' POSIX compatible.Gravatar Jasper Siepkes2018-04-09
| | | | | | | | | | The 'd_type' field is not part of the POSIX specification. Added a compile time check to see if we can use it. When not present fallback to a (slightly more expensive) call to 'stat'. This PR is similar to my other PR #4967. It's part of an effort to port Bazel to an POSIX compliant platform. Even though my porting effort may fail I think POSIX compliance could be beneficiary to Bazel either way. Closes #4969. PiperOrigin-RevId: 192096587
* Fixes a broken @link.Gravatar shahan2018-04-09
| | | | | | @link doesn't support generics, so this is converted instead to @code. PiperOrigin-RevId: 192094248
* Remove hashCode int field from Node.Gravatar dbabkin2018-04-09
| | | | | | | This is needless of complexity. Having counter in DiGraph can provide us with deterministic hashCode for every application run. There is no visible usage of this feature right now. More than, user of the DiGraph class should not rely on the fact the hashCode of the object will always be deterministic. Default Object.hashCode implementation returns deterministic result in cope of single JVM, and this is enough. RELNOTES:none PiperOrigin-RevId: 192093877
* Serialize Artifacts by their root-relative path rather than their exec path. ↵Gravatar janakr2018-04-08
| | | | | | The exec path contains redundant information when considered with the ArtifactRoot, so, since the ArtifactRoot is memoized, we can save space and time by only serializing the non-redundant root-relative path part. PiperOrigin-RevId: 192076469
* Cleanup Artifact.getPath() usage in AndroidDevice.Gravatar shahan2018-04-06
| | | | PiperOrigin-RevId: 191972075
* Cleanup Artifact.getPath() usage in Runfiles.Gravatar shahan2018-04-06
| | | | PiperOrigin-RevId: 191971225
* Improve golden testsGravatar cushon2018-04-06
| | | | PiperOrigin-RevId: 191970818
* In inlined skylark import lookup function caching, don't actually fetch deps ↵Gravatar shreyax2018-04-06
| | | | | | when all we want to do is register an edge and don't require the value. PiperOrigin-RevId: 191966203
* Emit a usage error if --output is not providedGravatar cushon2018-04-06
| | | | PiperOrigin-RevId: 191961686
* Flip default value of --experimental_shortened_obj_file_path to trueGravatar pcloudy2018-04-06
| | | | | | RELNOTES: Flip default value of --experimental_shortened_obj_file_path to true, Bazel now generates short object file path by default. PiperOrigin-RevId: 191958480
* Adds more debug logging to ArtifactFactory validation.Gravatar shahan2018-04-06
| | | | | | Blacklists vfs for DynamicCodec PiperOrigin-RevId: 191951230
* fix deps checker tool handling of primitive arrays in annotationsGravatar kmb2018-04-06
| | | | PiperOrigin-RevId: 191948995
* Documentation typo fixGravatar gregce2018-04-06
| | | | PiperOrigin-RevId: 191930771
* Cleanup @SkylarkCallable parameters and their error handling.Gravatar cparsons2018-04-06
| | | | | | | This involves enforcing additional compiletime restrictions on Param ordering and semantics. RELNOTES: None. PiperOrigin-RevId: 191927206
* Stop passing location of testdata as jvm property.Gravatar ajmichael2018-04-06
| | | | | | | We can use the builtin env variables to get this. RELNOTES: None PiperOrigin-RevId: 191926221
* Remove die() and replace it with BAZEL_DIE, part of the logging framework.Gravatar ccalvarin2018-04-06
| | | | | | | This will mean the messages will make it to the right output stream. RELNOTES: PiperOrigin-RevId: 191925662
* DynamicCodec emits a trail of type names when it encounters ↵Gravatar shahan2018-04-06
| | | | | | | | NoSuchCodecException. This will make it easier to trace down missing codecs. PiperOrigin-RevId: 191920743
* Cleanup use of Artifact.getPath() in AndroidResources.Gravatar shahan2018-04-06
| | | | PiperOrigin-RevId: 191918594
* Delete dead codeGravatar carmi2018-04-06
| | | | | RELNOTES: None PiperOrigin-RevId: 191916828
* Cleanup Artifact.getPath() usage from ResourceFilterFactoryGravatar shahan2018-04-06
| | | | PiperOrigin-RevId: 191914805
* Disallow undefined configs by default.Gravatar ccalvarin2018-04-06
| | | | | | | | This brings it in line with other flags, where a malformed value is an error. RELNOTES(INC): --config=foo will error if no 'foo' config exists. PiperOrigin-RevId: 191910781
* Add a new test environment variable, TEST_TARGET.Gravatar John Cater2018-04-06
| | | | | | | | | | | | | | | | | | | | Also include the test target in the test log for easier debugging of tests. Fixes #4960. Change-Id: Ic79bf21e8eeba765b607755c1d290ee4e910d01f Closes #4972. Sample output: ``` exec ${PAGER:-/usr/bin/less} "$0" || exit 1 Executing tests from //src/test/shell/bazel:bazel_example_test ----------------------------------------------------------------------------- ``` Change-Id: I433cd9251a564b269da5566bb8401151ae4c99b7 PiperOrigin-RevId: 191891278
* Automated rollback of commit 77c5582fa0f5c829df576d9f8d66f4975fe415a6.Gravatar hlopko2018-04-06
| | | | PiperOrigin-RevId: 191880445
* Prepare the file system before executing tests with --direct_run.Gravatar lberki2018-04-06
| | | | | RELNOTES: None. PiperOrigin-RevId: 191876954
* C++: Introduce provider that wraps C++ compilation providers.Gravatar plf2018-04-06
| | | | | | | For now, only CcCompilationContextInfo is wrapped. CcCompilationContextInfo will be renamed CcCompilationContext since the *Info suffix is used for providers. This will be done in a follow-up CL. RELNOTES:none PiperOrigin-RevId: 191876504
* Ensure arguments for repository rules are available earlyGravatar Klaus Aehlig2018-04-06
| | | | | | | | | | | ...so that a repository rule can access them in any order without being restarted. Restarting a repository rule can be expensive, if it already accessed the network or executed expensive commands. Improves on #4533. Change-Id: I1d27f88492e4e51fb955f3f066a28a75c5016169 PiperOrigin-RevId: 191874363
* Move shell executable to its own configuration fragment.Gravatar lberki2018-04-06
| | | | | RELNOTES: None. PiperOrigin-RevId: 191861074
* Properly report completion of shared actions with input discoveryGravatar hlopko2018-04-05
| | | | | | | | | | | | Currently we report "Analyzing" when include scanning runs. But since we can have shared C++ compile actions, only one of the group will be executed and only one will be reported completed. Remaining shared actions currently stay with "analyzing" forever. This cl makes sure that these actions are properly handled when finished. RELNOTES: None PiperOrigin-RevId: 191849728
* Make DynamicCodec always memoize. Since it replaces Java serialization, ↵Gravatar janakr2018-04-05
| | | | | | which memoizes, we should too. PiperOrigin-RevId: 191813677
* Tolerate empty bootclasspaths in javac-turbineGravatar cushon2018-04-05
| | | | PiperOrigin-RevId: 191809494
* Split Args#add into three methods.Gravatar tomlu2018-04-05
| | | | | | | | | | | Args#add(value, *, arg, format) Args#add_all(value, *, arg, map_each, format_each, before_each, omit_if_empty, uniquify) Args#add_joined(value, *, arg, join_with, map_each, format_each, format_joined, omit_if_empty, uniquify) The old Args#add remains backwards compatible, but we add a flag to disable this compatibility mode. RELNOTES: None PiperOrigin-RevId: 191804482
* Cleanup Artifact.getPath() call from ActionExecutionException.Gravatar shahan2018-04-05
| | | | PiperOrigin-RevId: 191801369
* Use @AutoCodec for NestedSet serialization.Gravatar janakr2018-04-05
| | | | PiperOrigin-RevId: 191797413
* Remove legacy manifest merging from Bazel.Gravatar ajmichael2018-04-05
| | | | | RELNOTES: android_binary.manifest_merger is no longer supported. PiperOrigin-RevId: 191791177
* Transform StackOverflowError thrown in ArrayCodec into ↵Gravatar janakr2018-04-05
| | | | | | SerializationException. This can happen with deep nested sets. PiperOrigin-RevId: 191790431
* Fix check-then-act race for some function potentially returning multiple ↵Gravatar shreyax2018-04-05
| | | | | | targets. PiperOrigin-RevId: 191774693
* Cleanup unneeded Artifact.getPath() call in ResourceFilterFactoryGravatar shahan2018-04-05
| | | | PiperOrigin-RevId: 191774614
* Introduces a SourceArtifact type.Gravatar shahan2018-04-05
| | | | | | | SourceArtifact.getPath() is considered safe and eventually we may restrict getPath() to only SourceArtifact. PiperOrigin-RevId: 191768519
* Make statically linking cpp-runtimes as a crosstool feature.Gravatar Googler2018-04-05
| | | | | | | | | | | This cl should result in no user-visible change. It simply sets the feature based on the field in the crosstool proto. A subsequent cl will expose this feature for people to actually use. DELTA_BY_EXTENSION=java=100,py=15 RELNOTES: None PiperOrigin-RevId: 191761838
* Error out gracefully instead of crashing blaze when --output=transitions is ↵Gravatar juliexxia2018-04-05
| | | | | | used without the --transitions flag PiperOrigin-RevId: 191755762
* Fix strict module map propagation for j2objc.Gravatar allevato2018-04-05
| | | | | | J2Objc needs special consideration to implement strict Swift-ObjC deps. The module maps for the generated Objective-C code are associated with their originating java_library targets (via the J2Objc aspect), which a swift_library cannot directly depend on. So those deeper down module maps need to be propagated transitively *until* a j2objc_library is reached; at that point, the j2objc_library should propagate all of its Java deps module maps, but *not* any of its j2objc_library deps (because that would be non-strict). PiperOrigin-RevId: 191754811
* Expose build file label retrieval as a protected member.Gravatar shreyax2018-04-05
| | | | PiperOrigin-RevId: 191752718
* Refactor CcToolchainFeatures$Variables to not create a map in the case of a ↵Gravatar janakr2018-04-05
| | | | | | singleton. PiperOrigin-RevId: 191744943
* Avoids calling getPath() in Artifact.toDetailString().Gravatar shahan2018-04-05
| | | | | | | | Also deletes obsolete Artifact.serializeToString() method. "Semantic" change: replaces the actual exec root with <execution_root> in the toString() representation. PiperOrigin-RevId: 191742669
* Fix the definition of Argument in the Skylark specificationGravatar vladmos2018-04-05
| | | | PiperOrigin-RevId: 191717115
* Remove c++-link-interface-dynamic-library from crosstools, it's not usedGravatar hlopko2018-04-05
| | | | | RELNOTES: None. PiperOrigin-RevId: 191713379
* Make code much easier to read without misleading variable name and negation.Gravatar twerth2018-04-05
| | | | | RELNOTES: None PiperOrigin-RevId: 191709934