| Commit message (Collapse) | Author | Age |
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144494739
MOS_MIGRATED_REVID=144494739
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144474876
MOS_MIGRATED_REVID=144474876
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is the preferred way to test for membership in, or iterate over, depsets (aka nested sets, aka plain old set()). The old way of doing membership tests or iterations over the raw depset itself is deprecated and may be removed in the future.
Note that membership testing in a depset was always an O(n) operation, perhaps contrary to the user's expectation, so using to_list() does not make things asymptotically worse. It just makes things more explicit.
RELNOTES: To iterate over or test for membership in a set, prefer using the new to_list() method. E.g., "for x in myset.to_list():", or "print(x in myset.to_list())". Iteration/membership-test on the raw set itself is deprecated.
--
PiperOrigin-RevId: 144452510
MOS_MIGRATED_REVID=144452510
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make it part of the interface contract that the
argv[0] must be either absolute or just a file
name. The only caller has been adhering to this
contract already, but document it explicitly
still.
--
PiperOrigin-RevId: 144447641
MOS_MIGRATED_REVID=144447641
|
|
|
|
|
|
|
|
| |
a transition and is completely unused by now.
--
PiperOrigin-RevId: 144446659
MOS_MIGRATED_REVID=144446659
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144444493
MOS_MIGRATED_REVID=144444493
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Breaks Bazel on JDK 7: http://ci.bazel.io/view/Dashboard/job/bazel-tests/BAZEL_VERSION=HEAD-jdk7,PLATFORM_NAME=linux-x86_64/475/console
*** Original change description ***
Automated [] rollback of commit 5163ec2c0db8fbdc7f367bb4aca5b6d7ae6d80b0.
*** Reason for rollback ***
See unknown commit
*** Original change description ***
Automated [] rollback of commit c9dbd4c1a4a4b4345c0b3924f86e01612ae12807.
*** Reason for rollback ***
Breaks bazel-docker-tests
http://ci.bazel.io/view/Dashboard/job/bazel-docker-tests/lastCompletedBuild/BAZEL_VERSION=HEAD,PLATFORM_NAME=docker/console
*** Original change description ***
Enable header compilation by default
--
PiperOrigin-RevId: 144443446
MOS_MIGRATED_REVID=144443446
|
|
|
|
|
|
|
|
|
| |
This should be a no-op change, primarily intended to improve the BlazeModule
API. The code simplification in the distributor code path is incidental.
--
PiperOrigin-RevId: 144441458
MOS_MIGRATED_REVID=144441458
|
|
|
|
|
|
|
|
|
|
|
|
| |
RUNFILES_DIR environment variable.
Note that this is slightly different from the Java launcher (where the same is JAVA_RUNFILES)
Fixes #2339.
--
PiperOrigin-RevId: 144439111
MOS_MIGRATED_REVID=144439111
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144434688
MOS_MIGRATED_REVID=144434688
|
|
|
|
|
|
|
|
| |
RedirectChaser needs to load the target to check if it's a bind/alias rule so it's not possible that it's not loaded yet by the time we get there.
--
PiperOrigin-RevId: 144429298
MOS_MIGRATED_REVID=144429298
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Technically, this code is incorrect, because we're calling into spawn strategy,
which might do any amount of additional waiting (e.g., to acquire local
resources). We should change spawn strategy to return structured information
about the subprocess execution, including the wall time, and use that to set
the test information. But for now, this seems good enough.
Fixes #1748.
Effectively a rollback of the rollback, but with a manual fix.
The problem was that the code was assuming exactly one process time,
which is incorrect in at least the case where the proto comes from a previous
Bazel version, because that didn't set the process time at all. We're now
copying over all the values, which should work in all cases.
--
PiperOrigin-RevId: 144425125
MOS_MIGRATED_REVID=144425125
|
|
|
|
|
|
|
|
| |
possible the exception has information we've been suppressing.
--
PiperOrigin-RevId: 144405370
MOS_MIGRATED_REVID=144405370
|
|
|
|
|
|
|
|
| |
and then wrapping it in another nested set.
--
PiperOrigin-RevId: 144398091
MOS_MIGRATED_REVID=144398091
|
|
|
|
|
|
|
|
|
|
|
| |
that may contain objc source. This only works with dynamic configurations
turned on.
This ensures that objc dependencies of non-objc targets use the apple crosstool.
--
PiperOrigin-RevId: 144384483
MOS_MIGRATED_REVID=144384483
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Rollforward with fix for AarImportTest. The test failed previously due to a difference in the configuration because getConfiguredTarget("//a:bar") does not invoke the android split transition, so the artifact paths did not exactly match.
*** Original change description ***
Automated [] rollback of commit 7fc35ae6890f4130b4d99a6c5007dcee91c5a8cc.
*** Reason for rollback ***
Breaks //src/test/java/com/google/devtools/build/lib:bazel-rules-tests
http://ci.bazel.io/view/Dashboard/job/bazel-tests/BAZEL_VERSION=HEAD,PLATFORM_NAME=linux-x86_64/lastCompletedBuild/console
*** Original change description ***
aar_import provides JavaCompilationArgsProvider for DexArchiveAspect.
--
PiperOrigin-RevId: 144370584
MOS_MIGRATED_REVID=144370584
|
|
|
|
|
|
|
|
| |
shouldn't come up very often, except in the linked bug.
--
PiperOrigin-RevId: 144367142
MOS_MIGRATED_REVID=144367142
|
|
|
|
|
|
|
|
| |
useful, but may also help identify Bazel as the culprit in the linked bug if we're lucky and the timestamps are clear.
--
PiperOrigin-RevId: 144359628
MOS_MIGRATED_REVID=144359628
|
|
|
|
|
|
|
|
| |
compilation actions.
--
PiperOrigin-RevId: 144356912
MOS_MIGRATED_REVID=144356912
|
|
|
|
|
|
|
|
| |
Moved some tests, fixed formatting, changed to use assertThat().
--
PiperOrigin-RevId: 144356402
MOS_MIGRATED_REVID=144356402
|
|
|
|
|
|
|
|
|
| |
(as they are for objc_library). This allows for swift/objc interop with
experimental_objc_library, since swift imports require compiled headers.
--
PiperOrigin-RevId: 144353367
MOS_MIGRATED_REVID=144353367
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144353268
MOS_MIGRATED_REVID=144353268
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
available to the extension registry generator.
With deploy_env, the message being extended may end up in the deploy_env binary's generated extension registry. However, the extension referencing that message is outside of deploy_env. For this case in the full runtime, we can synthesize the message as we don't need the loadPath (which is only used by lite runtime).
Rollback of commit 34de5d7d4a0678b3a4097cb7531fedb3ddb45983.
*** Reason for rollback ***
Rolling forward
*** Original change description ***
Automated [] rollback of commit fa7e34979b1a10794321b1233e8d6f44a62431fa.
--
PiperOrigin-RevId: 144346260
MOS_MIGRATED_REVID=144346260
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
See unknown commit
*** Original change description ***
Automated [] rollback of commit c9dbd4c1a4a4b4345c0b3924f86e01612ae12807.
*** Reason for rollback ***
Breaks bazel-docker-tests
http://ci.bazel.io/view/Dashboard/job/bazel-docker-tests/lastCompletedBuild/BAZEL_VERSION=HEAD,PLATFORM_NAME=docker/console
*** Original change description ***
Enable header compilation by default
--
PiperOrigin-RevId: 144344329
MOS_MIGRATED_REVID=144344329
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144339611
MOS_MIGRATED_REVID=144339611
|
|
|
|
|
|
|
|
| |
Fixes #1248.
--
PiperOrigin-RevId: 144333530
MOS_MIGRATED_REVID=144333530
|
|
|
|
|
|
|
|
| |
RELNOTES[INC]: All executable labels must also have a cfg parameter specified.
--
PiperOrigin-RevId: 144332992
MOS_MIGRATED_REVID=144332992
|
|
|
|
|
|
|
|
| |
Add Order#isCompatible, clean javadoc. Also fix broken @Deprecated-based canary for detecting accidental flattening.
--
PiperOrigin-RevId: 144331341
MOS_MIGRATED_REVID=144331341
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144325333
MOS_MIGRATED_REVID=144325333
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144323833
MOS_MIGRATED_REVID=144323833
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Breaks //src/test/java/com/google/devtools/build/lib:bazel-rules-tests
http://ci.bazel.io/view/Dashboard/job/bazel-tests/BAZEL_VERSION=HEAD,PLATFORM_NAME=linux-x86_64/lastCompletedBuild/console
*** Original change description ***
aar_import provides JavaCompilationArgsProvider for DexArchiveAspect.
--
PiperOrigin-RevId: 144316524
MOS_MIGRATED_REVID=144316524
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a separate argument to nativeCreateProcess for
argv[0] specifically, and another for the rest of
the args.
In a subsequent change I'll add code to compute
the 8dot3 style short name of the argv[0] so we
can use longer paths for executables in
CreateProcessA than we normally could. This is the
same approach as used in commit 44ecf9a0c7c25496a43f59f1c8f20df9527e12cb
See https://github.com/bazelbuild/bazel/issues/2107
See https://github.com/bazelbuild/bazel/issues/2181
--
PiperOrigin-RevId: 144311562
MOS_MIGRATED_REVID=144311562
|
|
|
|
|
|
|
|
| |
specify jar files containing Java resources.
--
PiperOrigin-RevId: 144299119
MOS_MIGRATED_REVID=144299119
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
1. no flag: only the direct reason is given (command execution termination status), and also the instruction to use "--verbose_failures"
2. flag "--verbose_failures": gives failed execution command and the instruction to use "--sandbox_debug --strategy"
3. flag "--verbose_failures --sandbox_debug": gives failed execution command, debugging message from sandboxing, and the instruction to use "--strategy"
Also removes "cd <sandbox_path>" in given failed command, since debugging is only necessary with flag "--verbose_failures --sandbox_debug" and the path is already given in sandboxing debugging message.
Addresses #2174.
Fixes bazel_sandboxing_test.sh and bazel_test_test.sh for the new error message.
Fixes timeout.
--
PiperOrigin-RevId: 144285435
MOS_MIGRATED_REVID=144285435
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
api_level is no longer required. It now sets the default android_sdk that is used if the --android_sdk flag is not passed. If it is not set, the highest api_level will be used as the default.
The new behavior is demonstrated by the following example:
$ ls $ANDROID_HOME/platforms
android-21 android-22 android-23
$ cat WORKSPACE
android_sdk_repository(
name = "foo",
)
$ bazel build //java/my/app # uses api level 23
$ bazel build --android_sdk=@foo//:sdk-22 //java/my/app # uses api level 22
$ cat > WORKSPACE <<EOF
android_sdk_repository(
name = "foo",
api_level = 21,
)
EOF
$ bazel build //java/my/app # uses api level 21
$ bazel build --android_sdk=@foo//:sdk-23 //java/my/app # uses api level 23
See https://github.com/bazelbuild/bazel/issues/2284 for the master plan for android_sdk_repository.
RELNOTES: android_sdk_repository no longer requires api_level. If one is not specified, the highest android platform installed will be used. Furthermore, android_sdk's are created for all android platforms installed and can be specified with the --android_sdk flag.
--
PiperOrigin-RevId: 144258881
MOS_MIGRATED_REVID=144258881
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144258810
MOS_MIGRATED_REVID=144258810
|
|
|
|
|
|
|
|
|
|
| |
incoming edges if dynamic configurations are turned on.
This CL does not cover top-level nodes.
--
PiperOrigin-RevId: 144258789
MOS_MIGRATED_REVID=144258789
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144257691
MOS_MIGRATED_REVID=144257691
|
|
|
|
|
|
|
|
|
|
| |
WindowsJniLoader.loadJniForTesting is just a
special case of what WindowsJniLoader.loadJni
already does, so we can just use the latter.
--
PiperOrigin-RevId: 144224388
MOS_MIGRATED_REVID=144224388
|
|
|
|
|
|
|
|
|
|
| |
--experimental_prune_more_modules. Both flags have proven to be stable for a
few weeks. Instead provide a (probably permanent) option to disable modules
pruning completely.
--
PiperOrigin-RevId: 144215146
MOS_MIGRATED_REVID=144215146
|
|
|
|
|
|
|
|
| |
* This new configuration allows for passthrough of Swift-specific options into Skylark for further consumption by swift_library. It also keeps things nicely isolated until we get a way to create configurations in Skylark directly.
--
PiperOrigin-RevId: 144210267
MOS_MIGRATED_REVID=144210267
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144207746
MOS_MIGRATED_REVID=144207746
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144207428
MOS_MIGRATED_REVID=144207428
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144194956
MOS_MIGRATED_REVID=144194956
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If linkstatic was explicitly set to 0 for a non-test target, we didn't set
CcLibraryHelper to create CcLinkOutputs, but would still try to link those in
instead of the compile output. Instead, pull out a variable that puts this
logic into a single spot.
Also rename the flag to --experimental_link_compile_output_separately, which
IMO makes slightly more sense. Not too important as I don't think we should
keep this flag long-term anyway.
--
PiperOrigin-RevId: 144194903
MOS_MIGRATED_REVID=144194903
|
|
|
|
|
|
|
|
|
|
| |
agent through the $jacocorunner dependency.
They are either both linked in or neither of them is, so there isn't much point in separating them.
--
PiperOrigin-RevId: 144191570
MOS_MIGRATED_REVID=144191570
|
|
|
|
|
|
|
|
|
|
| |
This is part of a series of changes to clean up the test strategies; this
specific change is a Google-internal refactoring. I'm planning to also update
the external test strategies subsequently.
--
PiperOrigin-RevId: 144184752
MOS_MIGRATED_REVID=144184752
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144181101
MOS_MIGRATED_REVID=144181101
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Breaks bazel-docker-tests
http://ci.bazel.io/view/Dashboard/job/bazel-docker-tests/lastCompletedBuild/BAZEL_VERSION=HEAD,PLATFORM_NAME=docker/console
*** Original change description ***
Enable header compilation by default
--
PiperOrigin-RevId: 144176396
MOS_MIGRATED_REVID=144176396
|
|
|
|
|
|
| |
--
PiperOrigin-RevId: 144127414
MOS_MIGRATED_REVID=144127414
|