| Commit message (Collapse) | Author | Age |
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121524513
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121520005
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, bazel would be internally inconsistent regarding whether it thought
module files were in use for LIPO builds. The merged CppCompilationContext
would not list .pcm files as inputs, because its useHeaderModules flag was
forced to false, but the CppModel would expand %module_file to a list of
.pcm files, because it thought that modules were in use. This results in
LIPO build failures due to missing files.
RELNOTES:
Fix interaction between LIPO builds and C++ header modules.
--
MOS_MIGRATED_REVID=121513092
|
|
|
|
|
|
|
|
|
| |
* Merge conflicts are now bright red warnings, as the previous merger was simultaneously stricter and looser than expected.
* Legacy resource rules are now detected and cause bright red warnings.
* Merge conflicts will test for equality either of file contents or parsed xml
--
MOS_MIGRATED_REVID=121510152
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121502983
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121497233
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* SimpleXmlResourceValue records all attributes except name. (I'd like to toss description, just as soon as I know I can -- bloats the serialized form.)
* Added handling for <item> to SimpleXmlResourceValue: this can be used as placeholders, or (apparently) mostly unvalidated value definitions.
* drawable can be used to put in colored backgrounds in a values.xml
* Strings, Plurals, need to read sub tags such as xliff as string
* <eat-comment/> and <skip/> need to be handled: we ignore them as all comments are eaten right now; and skip appears to be for localization diffs
* xml parser must be resilient to random characters in resources defined in values.
* Handles the <resource> not <resources> coding mistake gracefully.
* Supports a <public> declaration. This feature is more wildly used than the nonexistant documentation suggests.
--
MOS_MIGRATED_REVID=121490019
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Seems to have broken exoblaze builds. https://paste.googleplex.com/5185376662585344 happens when rebuilding after a trivial change (removed a single method).
*** Original change description ***
Rollforward of "Workaround for duplicate input basename bug in apple's libtool", with regex fix
--
MOS_MIGRATED_REVID=121483994
|
|
|
|
|
|
|
|
|
|
| |
* Uses the png cruncher for crunching pngs.
* Respects qualifiers when generating the values/values.xml.
* Improvements to testability.
* Standardizes on throwing merging exceptions instead of random IOExceptions.
--
MOS_MIGRATED_REVID=121483439
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some had incorrect words, or were missing a word and the end period.
Make it match with the descriptions in CommonCommandOptions.java.
Add a short documentation comment to |restart_reason| field.
--
Change-Id: Iac802eb3ed4f72dfd402e85d2098a5616b40a9de
Reviewed-on: https://bazel-review.googlesource.com/3530
MOS_MIGRATED_REVID=121478703
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121475668
|
|
|
|
|
|
|
|
|
| |
prefixes" and compiler plugins.
They were dead code.
--
MOS_MIGRATED_REVID=121473393
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121472172
|
|
|
|
|
|
|
|
|
| |
doesn't exist.
It may be an old version that doesn't know how to write them yet.
--
MOS_MIGRATED_REVID=121462577
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, bazel_tools would be under whatever.runfiles/workspace-name/external/bazel_tools. However, we allowed "" as the workspace name (and Bazel's workspace name was "") so a lot of scripts started assuming that bazel_tools would be at whatever.runfiles/external/bazel_tools.
Going forward, the runfiles tree looks like:
whatever.runfiles/
__main__/
...
bazel_tools/
tools/
objc/
...
some_other_repo/
...
So all of the repos are on the same "level" of the directory tree. This lets you refer to runfiles the same way (x.runfiles/repo/path/to/your/thing) whether you're running locally or importing it as an external repository (see https://github.com/bazelbuild/bazel/wiki/Updating-the-runfiles-tree-structure for more detail than you probably care about).
Fixes #1227.
--
MOS_MIGRATED_REVID=121435276
|
|
|
|
|
|
|
| |
ApplicationManifest#getOverridePackage for building split and stub manifests.
--
MOS_MIGRATED_REVID=121426669
|
|
|
|
|
|
|
|
|
|
|
| |
cc_toolchain_suite. This will allow selecting the correct toolchain based on
--compiler / --android_compiler.
RELNOTES: The key for the map to cc_toolchain_suite.toolchains is now a string
of the form "cpu|compiler" (previously, it was just "cpu").
--
MOS_MIGRATED_REVID=121418076
|
|
|
|
|
|
|
|
|
| |
* values directories now respect qualifiers (whoops.)
* qualifiers must deal with 3-4 letter region codes
* qualifiers must add api version for normalization
--
MOS_MIGRATED_REVID=121417370
|
|
|
|
|
|
|
| |
parse resource file to use UTF_8 explicitly. Fixes crash when reading R.txt with non-ASCII characters.
--
MOS_MIGRATED_REVID=121415056
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As select statement requires to have all dependencies
resolved, we need to have all tooling of objc target
when doing a select if such targets are in the non
selected condition. This will enable to build
conditionally iOS targets only if we are on OS X.
This will permit to remove one of our hack from the
build.sh script of the tutorial.
--
MOS_MIGRATED_REVID=121411892
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Context:
AndroidLibrary provides JavaSourceInfoProvider, which JackAspect wants.
But it doesn't advertise it, so it will never have JackAspect applied.
This is good, because they both have an :android_sdk attribute and
want to provide JackLibraryProvider.
Rules in the external repository created with bind() don't have their
real advertisements, but instead are considered to advertise everything.
So, when an android_library depends on another android_library through
a bind() alias...
Boom.
This avoids the boom by making JackAspect bail out immediately if its
current target looks like an Android rule. It wouldn't be doing anything
useful anyway if the base rule already provides Jack libraries.
Fixes #1222.
--
MOS_MIGRATED_REVID=121401981
|
|
|
|
|
|
|
| |
ConfiguredTargetTransitivePackagesTest.
--
MOS_MIGRATED_REVID=121398365
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121397798
|
|
|
|
|
|
|
| |
and use it to validate that :java_toolchain has a JavaToolchainProvider.
--
MOS_MIGRATED_REVID=121396726
|
|
|
|
|
|
|
|
|
| |
package's package root when creating top-level directories. Otherwise, if the empty package references files in those top-level directories, they would be hidden.
Fixes #1221
--
MOS_MIGRATED_REVID=121392128
|
|
|
|
|
|
|
| |
can be added to by fragments. Uses the registry to enable native access to TranstiveProviderInstances from Skylark. In particular, makes ObjcProvider accessible to skylark rule implementations.
--
MOS_MIGRATED_REVID=121390911
|
|
|
|
|
|
|
| |
libtool", with regex fix
--
MOS_MIGRATED_REVID=121390817
|
|
|
|
|
|
|
| |
Serializer no longer serializes transitive resources. That turned out to be a performance issue with large trees. Instead,it relies on the build tool to indicate the direct and transitive resources. This reduces the number of keys that get pulled from diamond dependencies and stays much, much leaner.
--
MOS_MIGRATED_REVID=121382705
|
|
|
|
|
|
|
| |
"abi" / "abi_deps" are being removed imminently.
--
MOS_MIGRATED_REVID=121380507
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes https://github.com/bazelbuild/bazel/issues/990.
It's sometimes useful to be able to access the file pulled down
with maven_jar; for example, scala macro jars don't work with ijar,
so it's necessary to pull them into the compile-time classpath
unharmed.
--
Change-Id: I1d9e1cb30ea6fa56c18cbfcc68ee9b203524dfd4
Reviewed-on: https://bazel-review.googlesource.com/#/c/3550
MOS_MIGRATED_REVID=121379551
|
|
|
|
|
|
|
|
|
| |
InfoItems can
now take parameters.
--
MOS_MIGRATED_REVID=121379097
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
See https://github.com/bazelbuild/bazel/wiki/Updating-the-runfiles-tree-structure
for more details about this change.
RELNOTES[INC]: Bazel will now default to using __main__ as the workspace name
(instead of "", as previously) if a project does not provide a workspace
name in the WORKSPACE file. The repository's runfiles will appear under
x.runfiles/__main__/.
--
MOS_MIGRATED_REVID=121376094
|
|
|
|
| |
MOS_MIGRATED_REVID=121373163
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Breaks backwards compatibility.
*** Original change description ***
Collect logic related to extra Proguard outputs in one place.
--
MOS_MIGRATED_REVID=121363782
|
|
|
|
| |
MOS_MIGRATED_REVID=121353990
|
|
|
|
|
|
|
|
|
|
|
| |
I believe the original intention of this error check was to always fail if kwargs were present (while Skylark doesn't support kwargs, at least). However, the current code will most likely fail at findJavaMethod because only the args (i.e. not the kwargs) are used to find a matching method.
This issue has already been encountered on the bazel-discuss group (g/bazel-discuss/hHyf2txlS70).
With the proposed modification, I believe any kwargs provided to Skylark methods will cause the kwarg-specific error.
--
MOS_MIGRATED_REVID=121345719
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes #1166.
While I was here, I also:
* Renamed the implementations of Decompressor to show that they are
Decompressors, not SkyFunctions (they used to be, it's confusing to have them
still have the -Function suffix).
* Added a unit test and moved the testing target to the rule/repository
subdirectory, since it was a good chance to break off a tiny chunk of the
monolithic BUILD file in lib/.
--
MOS_MIGRATED_REVID=121306165
|
|
|
|
|
|
|
|
| |
* This flag forces -g and enables dSYM bundle actions.
* All instances of breakpad generation are now gated by the old --objc_generate_debug_symbols option.
--
MOS_MIGRATED_REVID=121304681
|
|
|
|
|
|
|
|
|
| |
These attributes are going away imminently.
select() is the new abi_deps.
--
MOS_MIGRATED_REVID=121294764
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As of 18c277f, we have special logic for mixing
selects and constraints. With that logic, a multiplaform
libray declares all environments it supports across
all select paths, then each path supports some subset of
those environments.
With this approach, there's no need for
compatible_with = select(). So we'd like to disallow
that pattern (and all its downsides) in preference for
what 18c277f allows.
--
MOS_MIGRATED_REVID=121294433
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=121291329
|
|
|
|
|
|
|
| |
Creates a symlink for each input file to libtool, with a md5 hash suffix, and passes these symlinks instead.
--
MOS_MIGRATED_REVID=121279266
|
|
|
|
|
|
|
|
| |
* Adds a new method to get Xcode's developer frameworks dir.
* Removes platform_dir() as it cannot be used in Skylark as-is - there's no way to get the platform name it requires.
--
MOS_MIGRATED_REVID=121275439
|
|
|
|
|
|
|
|
|
|
|
| |
actually a server process.
This should be implemented for other OSes, too, but OS X seems to lack a procfs and it's not clear how to discover anything about a process based on its PID and of course, Windows is a wholly different cup of tea.
More work for #930.
--
MOS_MIGRATED_REVID=121262673
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The test_terminal_columns_default as part of the rc-options test incorrectly
assumed that (in the standard UI) the default width for terminal columns should
be long enough to not need line wrapping, even for a 32 character long
label. However, the default value is 80, and another 37 characters quickly come
together: 8 for the progress information ("[3 / 4] "), 23 for the activity
("Creating runfiles tree "), and, depending on the actual build, another couple
of characters for the the path fragment (like "bazel-out/local_linux-fastbuild",
which has 31 characters on its own). So remove that test based on wrong assumptions.
--
Change-Id: I6019f8d30e774ade0340bcec8f7ddb3768068135
Reviewed-on: https://bazel-review.googlesource.com/#/c/3541
MOS_MIGRATED_REVID=121262554
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Made the control flow much simpler and more understandable
- Added some documentation about the interplay of the client and the server
- Abstracted out POSIX mechanisms from blaze.cc so that they can be implemented properly on Windows
- Added assertions that the methods on BlazeServer are called when they should be
Polish for #930.
--
MOS_MIGRATED_REVID=121256601
|
|
|
|
|
|
|
| |
Python version priors to 2.5 doesn't support ternary conditional operator.
--
MOS_MIGRATED_REVID=121256017
|
|
|
|
|
|
|
| |
actually need it.
--
MOS_MIGRATED_REVID=121160209
|
|
|
|
|
|
|
|
|
|
|
| |
command to see if a flag policy applies, check whether the flag applies by seeing if its list of commands matches one of the commands in the hierarchy.
This avoids the tedious and brittle specification of all commands that may use a flag, while providing better filtering out of inapplicable flags.
RELNOTES: A FlagPolicy specified via the --invocation_policy flag will now match the current command if any of its commands matches any of the commands the current command inherits from, as opposed to just the current command.
--
MOS_MIGRATED_REVID=121159131
|
|
|
|
|
|
|
|
|
|
|
| |
assumed-to-be-immutable filesystem type of output_base.
When we do have an OutputService, defer computation of the OutputService's dynamic filesystem type until we actually need it at the start of a "build" or "test" command.
This saves ~450us on each invocation, according to my benchmarks.
--
MOS_MIGRATED_REVID=121156323
|