| Commit message (Collapse) | Author | Age |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Causing Bazel server to crash when running bazel clean --expunge
https://github.com/bazelbuild/bazel/issues/3956
*** Original change description ***
Delayed rollforward of commit 8fb311b4dced234b2f799c16c7d08148619f4087.
This was rolled back due to Tensorflow breakage but the patch I exported to gerrit (https://bazel-review.googlesource.com/c/bazel/+/18590) passed Tensorflow (https://ci.bazel.io/job/bazel/job/presubmit/52/Downstream_projects/). Confirmed with jcater@ that the "newly failing" projects in the Global Tests are known issues. I think we can check this in now.
Additionally I had attempted to reproduce any tensorflow issues with this by building and testing Tensor...
***
ROLLBACK_OF=172361085
RELNOTES:None
PiperOrigin-RevId: 175821671
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Even if an action is planned to run and a strategy assigned,
it does not mean it is actually executing; e.g., it might be
waiting for local resources. To reduce confusion, show the
status instead of the execution strategy for non-executing
actions. Also avoid the word "running" for the total number
of executing and scheduled to be executed actions.
Improves on Issue #4089.
Change-Id: If658c1a24ee26eb27ccd892847af18015355a8d3
PiperOrigin-RevId: 175818071
|
|
|
|
|
| |
Change-Id: If8f4c75b65d5654703a68fe70277dc597773c85d
PiperOrigin-RevId: 175737062
|
|
|
|
|
|
|
|
|
|
|
| |
does not depend on AutoCloseable, as it is not available before API 19.
This CL includes the following:
1. A type inference algorithm based on ASM. It relies on the stack map frames to compute type information at the entry of basic blocks.
2. The type inference is used to infer the types of the resources to be closed. Then for each concrete resource type, we specialize the synthetic $closeResource method to $closeResource(Throwable, <concrete resource type>).
RELNOTES: None
PiperOrigin-RevId: 175731437
|
|
|
|
|
|
| |
Closes #3972.
PiperOrigin-RevId: 175726041
|
|
|
|
| |
PiperOrigin-RevId: 175718465
|
|
|
|
| |
PiperOrigin-RevId: 175709097
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The deps are passed through into the generated java_import and aar_import rules. This is necessary for AARs with resource dependencies since maven_aar ignores transitive dependencies. It's less significant for the Java rules, since typically JARs on Maven are compiled class files and as such only have runtime dependencies.
Example usage:
```
# WORKSPACE
load("@bazel_tools//tools/build_defs/repo:maven_rules.bzl", "maven_aar")
maven_aar(
name = "android_image_cropper",
artifact = "com.theartofdev.edmodo:android-image-cropper:2.3.1",
deps = [
"@androidsdk//com.android.support:appcompat-v7-24.1.1",
],
)
```
Fixes https://github.com/bazelbuild/bazel/issues/2863.
Fixes https://github.com/bazelbuild/bazel/issues/3980.
Also, cleans up some broken stuff in the tests (sadly, they don't run on jenkins, so nothing caught that they were broken). test_maven_jar_with_classifier_skylark still does not work, so I disabled.
RELNOTES: Add deps attribute to Skylark maven_aar and maven_jar workspace rules.
PiperOrigin-RevId: 175698282
|
|
|
|
|
| |
RELNOTES: None
PiperOrigin-RevId: 175697300
|
|
|
|
|
|
|
|
|
| |
This is some clean up pulled out of unknown commit. It avoids needing
to reconstruct the dwoFile, which will be harder when we move to
using shared LTO backends in some cases.
RELNOTES: None
PiperOrigin-RevId: 175692708
|
|
|
|
| |
PiperOrigin-RevId: 175682806
|
|
|
|
|
|
|
|
|
|
|
| |
Quote empty args in the command argument vector
otherwise the command's executable would not
recognize them as arguments.
Fixes https://github.com/bazelbuild/bazel/issues/3973
Change-Id: Iccfb59e75c748247c8df7d52fb8cf4227eae6fa7
PiperOrigin-RevId: 175672201
|
|
|
|
|
|
|
|
| |
The wrapper script is now only necessary for TensorFlow GPU build.
Since the link option for MSVCRT is controlled by CROSSTOOL, we should remove the hack in wrapper script to avoid conflict.
RELNOTES: None
PiperOrigin-RevId: 175664202
|
|
|
|
|
|
|
|
| |
RELNOTES[NEW]: Users can use win_def_file attribute to specify a DEF file for
exporting symbols when build a shared library on Windows.
Change-Id: Ifa28d8b7b24eaefcefc9640d8dc56fd2931e9688
PiperOrigin-RevId: 175651203
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Breaks Google-internal targets, sadly.
*** Original change description ***
Ban middlemen from runfiles artifacts.
Previous changes have removed all middlemen from runfiles
artifacts. This CL locks it down and removes various now-redundant
*WithoutMiddlemen() methods from Runfiles.
I put a check for middlemen in ConflictChecker.put, which should be a
chokepoint for runfiles artifacts. It's unfortunate we can't detect
middlemen earlier than execution, but I can't see a way to efficiently
check every runfiles artifact earlier.
Cha...
***
RELNOTES: None.
PiperOrigin-RevId: 175650018
|
|
|
|
|
|
|
|
| |
For build log files that are inlined, we eventually have to
report ByteStrings, so avoid the detour through strings,
potentially causing encoding issues.
PiperOrigin-RevId: 175647576
|
|
|
|
| |
PiperOrigin-RevId: 175630707
|
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 175613518
|
|
|
|
|
|
|
| |
GITHUB: #903
RELNOTES: None.
PiperOrigin-RevId: 175600267
|
|
|
|
|
|
|
| |
incorrectly using multi-arch iOS
RELNOTES: None.
PiperOrigin-RevId: 175579695
|
|
|
|
|
| |
RELNOTES: none
PiperOrigin-RevId: 175576296
|
|
|
|
| |
PiperOrigin-RevId: 175558546
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Extend our integration test that verifies that bazel can
successfully be bootstrapped from the distribution artifact,
to also verify that SOURCE_DATE_EPOCH is honored. As the
bazel binary created contains a time stamp, we need to support
it being set from the invocation to ensure a reproducible
build. The standard way of setting time stamps to be embedded
is setting the SOURCE_DATE_EPOCH environment variable. So ensure
that that time stamp occurs in the version output.
Change-Id: Ib6388104d83426b6a74d995c324adee586560fe7
PiperOrigin-RevId: 175546459
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Mac OS X is the name of operating systems released by Apple, Inc between years 2001 and 2015. macOS is the name of operating systems released by Apple, Inc starting in 2016 and continuing until current times.
Documentation of the change in product name:
https://en.wikipedia.org/wiki/MacOS#macOS_10.12_Sierra
http://www.businessinsider.com/wwdc-2016-os-x-becomes-macos-2016-6
Closes #3990.
PiperOrigin-RevId: 175535106
|
|
|
|
| |
PiperOrigin-RevId: 175535010
|
|
|
|
|
|
| |
configuration.
PiperOrigin-RevId: 175532550
|
|
|
|
| |
PiperOrigin-RevId: 175531318
|
|
|
|
|
|
|
|
|
|
| |
runs_per_test allows multiple argument instances, because there's an
option of using a per-label regex format.
Unlike most such flags, earlier instances of the flag were dominating
later ones.
PiperOrigin-RevId: 175530675
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If you have java 9 installed but JAVA_HOME is set to java 8 then some bazel tests will fail because the check for java cannot find rt.jar which no longer exists in java 9.
e.g.
on OSX 10.13
```
> java -version
java version "1.8.0_144"
Java(TM) SE Runtime Environment (build 1.8.0_144-b01)
Java HotSpot(TM) 64-Bit Server VM (build 25.144-b01, mixed mode)
> JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_144.jdk/Contents/Home bazel test //src/test/py/bazel:action_temp_test
```
will fail with the following
```
FAIL: testLinuxOrDarwinSandboxedBuildAction (__main__.ActionTempTest)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/private/var/tmp/_bazel_cheister/08d5287ffe0da9d62c01970651a9400e/bazel-sandbox/7165594205799114739/execroot/io_bazel/bazel-out/local-fastbuild/bin/src/test/py/bazel/action_temp_test.runfiles/io_bazel/src/test/py/bazel/action_temp_test.py", line 30, in testLinuxOrDarwinSandboxedBuildAction
self.AssertTempEnvvarWithSpawnStrategy('sandboxed')
File "/private/var/tmp/_bazel_cheister/08d5287ffe0da9d62c01970651a9400e/bazel-sandbox/7165594205799114739/execroot/io_bazel/bazel-out/local-fastbuild/bin/src/test/py/bazel/action_temp_test.runfiles/io_bazel/src/test/py/bazel/action_temp_test.py", line 39, in AssertTempEnvvarWithSpawnStrategy
strategies = self._SpawnStrategies()
File "/private/var/tmp/_bazel_cheister/08d5287ffe0da9d62c01970651a9400e/bazel-sandbox/7165594205799114739/execroot/io_bazel/bazel-out/local-fastbuild/bin/src/test/py/bazel/action_temp_test.runfiles/io_bazel/src/test/py/bazel/action_temp_test.py", line 124, in _SpawnStrategies
self.AssertExitCode(exit_code, 2, stderr)
File "/private/var/tmp/_bazel_cheister/08d5287ffe0da9d62c01970651a9400e/bazel-sandbox/7165594205799114739/execroot/io_bazel/bazel-out/local-fastbuild/bin/src/test/py/bazel/action_temp_test.runfiles/io_bazel/src/test/py/bazel/test_base.py", line 68, in AssertExitCode
'(end stderr)------------------------------------------',
AssertionError: Bazel exited with 1 (expected 2), stderr:
(start stderr)----------------------------------------
INFO: $TEST_TMPDIR defined: output root default is '/private/var/tmp/_bazel_cheister/08d5287ffe0da9d62c01970651a9400e/bazel-sandbox/7165594205799114739/execroot/io_bazel/_tmp/30efe0ec7addffc49fb72b5d5aa45572' and max_idle_secs default is '15'.
Extracting Bazel installation...
Problem with java installation: couldn't find/access rt.jar in /Library/Java/JavaVirtualMachines/jdk-9.jdk/Contents/Home
(end stderr)------------------------------------------
```
This is related to https://github.com/bazelbuild/bazel/issues/428 but really only fixes the case where you are still using java8 but happen to have java9 installed.
Closes #4045.
PiperOrigin-RevId: 175513188
|
|
|
|
|
| |
RELNOTES: None.
PiperOrigin-RevId: 175508377
|
|
|
|
|
|
| |
This test is hitting time out on remote execution. Until the workers are fixed, deactivate it.
PiperOrigin-RevId: 175478669
|
|
|
|
|
|
| |
Issue bazelbuild/continuous-integration#149: FreeBSD 12 slaves cannot copy the Bazel binary keeping permission correctly.
PiperOrigin-RevId: 175474382
|
|
|
|
|
|
| |
Those tools are dependent on the host bazel (the one used for running the test) and that caused the test to be broken by the nightly.
PiperOrigin-RevId: 175462157
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
memory-saving non-incremental mode independent of --batch and --discard_analysis_cache.
A command run with --nokeep_incrementality_data will discard data that would be needed for incremental builds. Subsequent commands can be sent to the same server, but they will not get the benefit of incrementality from this command. However, if --keep_incrementality_data is specified on a subsequent command, the commands after that will get the benefits of incrementality.
There are two benefits to not being dependent on --batch. First, this allows Bazel servers to be run in extreme memory-saving mode without the startup penalties (JVM startup, JITting) that --batch execution imposes. Second, this allows Bazel developers to inspect the state of a Bazel server after an extreme memory-saving build.
In order to avoid discarding data unnecessarily (for instance, on a "bazel info used-heap-size-after-gc" or "bazel dump --skyframe=summary") the actual resetting of the graph is done lazily, right before its use in SequencedSkyframeExecutor#sync. This is morally a partial rollback of https://github.com/bazelbuild/bazel/commit/98cd82cbdcac7c48164a611c5a9aa8fc2f1720ef.
For now, our tests specify all of the flags. After this change sticks, I plan to get rid of the --batch flag from these tests, which should allow for some clean-ups. Eventually --batch and --discard_analysis_cache may not imply that we don't keep incremental state: we can require that it be specified explicitly.
PiperOrigin-RevId: 175335075
|
|
|
|
|
|
|
|
|
|
|
| |
While HTTP is faster and SHA256 already guaranteed data
hasn't been tampered with, it does give visibility over
the wire regarding which jars are being transmitted. The
true purpose of this change is to bring these practices
into greater consistency with Google's broader vision of
using HTTPS as much as possible.
PiperOrigin-RevId: 175328286
|
|
|
|
|
|
|
|
|
|
|
| |
It was added long ago, for reasons which are no longer relevant or
useful.
Also remove the build variable it controlled ("skip_mostly_static")
from CROSSTOOL files.
RELNOTES: None.
PiperOrigin-RevId: 175325225
|
|
|
|
|
|
|
|
|
|
| |
Also changed truth.SubjectFactory to truth.Subject.Factory (plain renaming) and use method reference instead of anonymous class to create the factory when applicable.
FailureMetadata, an opaque object to its users, is introduced to replace FailureStrategy in in custom Subject in order to resolve some existing flaws of FailureStrategy as well as enable new features to be added to Truth.
New API is available in Truth-0.36, if there is a build/pom.xml, it's also updated to use this version.
PiperOrigin-RevId: 175308179
|
|
|
|
|
|
| |
on a graph without edges. Now we fail gracefully.
PiperOrigin-RevId: 175294923
|
|
|
|
|
|
|
| |
Accept warnings specified by flag. This flag will cause a warning to be printed like a standard Option deprecation warning or any other warning from options parsing. This can be used to deprecate definitions of config expansion or bazelrc files. This should not be used to deprecate internal option specifications.
RELNOTES: None.
PiperOrigin-RevId: 175231963
|
|
|
|
| |
PiperOrigin-RevId: 175230893
|
|
|
|
|
|
| |
instead of assuming it is a deploy jar.
PiperOrigin-RevId: 175229774
|
|
|
|
|
|
|
| |
This feature is not yet available for Bazel users.
RELNOTES: n/a
PiperOrigin-RevId: 175226288
|
|
|
|
|
|
|
|
|
| |
constraint checking.
Fixes #4042.
Change-Id: I9461c004ab1aab8ca7a0ad515e8a9f24c8a489a5
PiperOrigin-RevId: 175219823
|
|
|
|
| |
PiperOrigin-RevId: 175218074
|
|
|
|
|
|
|
|
|
|
|
| |
Reorganize the BlazeOptionHandler to make it easy to exchange out how the config options are expanded using the definitions in the blazerc.
Also change the getOptionsMap call to actually structure the rc options in the order that we parse them: we expand them in command order (for the test command, first add all "common" options, then "build," then "test") and then within each command, we expand the options in the rc order. This somewhat simplifies the rc expansion code, and avoids the two-phase regrouping that used to happen.
This change should not change the semantics of rc option ordering. A followup change will add an alternative implementation for --config's expansion.
RELNOTES: None.
PiperOrigin-RevId: 175208971
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previous changes have removed all middlemen from runfiles
artifacts. This CL locks it down and removes various now-redundant
*WithoutMiddlemen() methods from Runfiles.
I put a check for middlemen in ConflictChecker.put, which should be a
chokepoint for runfiles artifacts. It's unfortunate we can't detect
middlemen earlier than execution, but I can't see a way to efficiently
check every runfiles artifact earlier.
Change-Id: I57fcb51b02e40adf3ebc4591c26bf5492c89352f
PiperOrigin-RevId: 175206157
|
|
|
|
| |
PiperOrigin-RevId: 175206100
|
|
|
|
|
|
|
|
|
|
| |
The build event protocol reports certain actions that are particularly
important to the user, in particular failed ones. Also report the precise
command line of those actions, if available. This is useful for IDEs
wanting to report errors in a more precise way.
Change-Id: I04224185ea46c608ad0dfcdb2e772071ca0736b3
PiperOrigin-RevId: 175203502
|
|
|
|
|
|
|
| |
This feature is experimental, and thus left mostly undocumented.
RELNOTES: None.
PiperOrigin-RevId: 175187791
|
|
|
|
|
|
|
|
| |
Add Apple Rules to index.
Closes #3992.
PiperOrigin-RevId: 175175081
|