| Commit message (Collapse) | Author | Age |
|
|
|
|
| |
--
MOS_MIGRATED_REVID=107726105
|
|
|
|
|
|
|
| |
is guaranteed not to be a package. Instead, start with its parent. This will be faster and take less memory.
--
MOS_MIGRATED_REVID=107725767
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=107719658
|
|
|
|
|
|
|
|
|
|
|
|
| |
Move
the trimming logic into key() itself to:
a) eliminate code redundancy
b) guarantee all future calls to key() also do this.
--
MOS_MIGRATED_REVID=107713353
|
|
|
|
|
|
|
| |
stack depth.
--
MOS_MIGRATED_REVID=107688035
|
|
|
|
|
|
|
| |
default this is disabled.
--
MOS_MIGRATED_REVID=107644420
|
|
|
|
|
|
|
|
|
| |
configuration transition *before* trimming the options. This
is because the transition may read fragments that aren't
in the dep itself.
--
MOS_MIGRATED_REVID=107624186
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
trim the BuildOptions to only those needed by the fragment.
This guarantees that the same fragment only gets instantiated
once (for example: we don't want to have two copies of a
CppConfiguration because one copy had PythonOptions and the
other didn't).
This also prevents crashes, e.g. in the Preconditions check in
CcLibraryHelper.addDeps.
--
MOS_MIGRATED_REVID=107609270
|
|
|
|
|
|
|
| |
Otherwise we have potential correctness and performance problems on a missing Skylark import dep.
--
MOS_MIGRATED_REVID=107605200
|
|
|
|
|
|
|
|
|
| |
replacing MemoizingEvaluator#getValues et al with WalkableGraph usage. I initially attempted to do this but punted once I realized how much work it would be.
Also make DirectoryListingStateValue and FileStateValue public for use in outside callers of FilesystemValueChecker.
--
MOS_MIGRATED_REVID=107447425
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Any Package name starting with a blacklisted prefix is treated as a deleted package.
Note the relation to prior art in Blaze:
1. Deleted packages indicates to Blaze to pretend certain packages do not exist. The deleted packages are passed in as fully resolved package names. This is important because the invalidation logic when these values changes references the exact package names, as opposed to prefix patterns.
2. Negative patterns in target parsing. (eg, "foo/..., -foo/bar/..."). Note that if //foo:biz depends on //foo/bar:bang, we still visit that dependency. In other words, the negative patterns just control the target pattern parsing, but not the later evaluation of those targets.
Blacklisting acts like both (1) and (2), but in the form of a Precomputed injected value (so modifying the value would be more expensive than modifying deleted_packages).
--
MOS_MIGRATED_REVID=107431993
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=107420673
|
|
|
|
|
|
|
|
| |
Aspect => ConfiguredAspect
AspectWithParameters => Aspect
--
MOS_MIGRATED_REVID=107375211
|
|
|
|
|
|
|
|
|
| |
The loading phase runner is now always created by the SkyframeExecutor, and
the duplicate test is dropped in favor of subclassing and overriding one
method.
--
MOS_MIGRATED_REVID=107188756
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
configurations get trimmed to the same fragments as target
configurations. This isn't correct, because null fragments
in the target configuration (which in practice don't exist)
shouldn't necessarily be excluded from the host config.
In the worst case then can crash fragment loaders, which
expect the fragment options to exist even if the fragment
ends up being null.
--
MOS_MIGRATED_REVID=107173093
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=107071400
|
|
|
|
|
|
|
| |
package loading from the LoadingCache default of 4 to value of --legacy_globbing_threads.
--
MOS_MIGRATED_REVID=107056166
|
|
|
|
|
|
|
|
|
| |
aspect-attributes.
The former contains both.
--
MOS_MIGRATED_REVID=106961145
|
|
|
|
|
|
|
|
| |
This makes the new loading phase runner work for basic builds, but there are
still a lot of broken tests if enabled by default.
--
MOS_MIGRATED_REVID=106952054
|
|
|
|
|
|
|
| |
More testing required before we even add a command-line option.
--
MOS_MIGRATED_REVID=106922350
|
|
|
|
|
|
|
|
| |
This is in preparation for adding a Skyframe-based implementation that uses
the recently added TargetPatternPhaseFunction.
--
MOS_MIGRATED_REVID=106920157
|
|
|
|
|
|
|
| |
Aspect provides.
--
MOS_MIGRATED_REVID=106882046
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=106848269
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=106836859
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=106694515
|
|
|
|
|
|
|
|
| |
There's no timeline right now, but it's unusable on OSX and Oracle has no
plans of fixing it.
--
MOS_MIGRATED_REVID=106486015
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=106444993
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
declare required configuration fragments: from
*everything* to *nothing*.
Now that all builtin rules properly declare their fragments,
the "backwards compatibility" concern that inspired the
original behavior is no longer needed.
This impacts, for example, filegroup rules, which have
nothing to declare.
--
MOS_MIGRATED_REVID=106433791
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=106323413
|
|
|
|
|
|
|
| |
maximum size, based off a reasonably larged sized benchmark.
--
MOS_MIGRATED_REVID=106180595
|
|
|
|
|
|
|
|
|
|
| |
Simplifies the runtime graph when TransitiveTraversalFunction is used.
Also moves an error reporting method from the base function to the
TransitiveTargetFunction, which is the only one that uses it.
--
MOS_MIGRATED_REVID=106109745
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105994907
|
|
|
|
|
|
|
| |
For native aspects, AspectClass is a facade for Class<AspectFactory<...>>.
--
MOS_MIGRATED_REVID=105986390
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, a TransitiveTraversalFunction computed a value that
contained an exception only if the exception resulted from a failure
to load the function's immediate target. If the target had transitive
dependencies on other targets, those other targets would be loaded, but
any errors resulting from loading those targets would not be retained.
SkyQueryEnvironment impromperly used a SkyFrame mechanism (which was
solely intended to ensure equivalent semantics between keep-going and
no-keep-going evaluations) to discover errors in the set of transitive
children of TransitiveTraversal nodes.
In order to transition SkyQueryEnvironment away from that mechanism,
this CL changes TransitiveTraversalFunction to remember the first error
message encountered while loading its target and its transitive
dependencies.
By remembering just the error message as a string, and not the full
exception object, this also helps TransitiveTraversalValue have more
consistent equality semantics for change-pruning.
--
MOS_MIGRATED_REVID=105977182
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105969015
|
|
|
|
|
|
|
| |
REPOSITORY_NAME symbol.
--
MOS_MIGRATED_REVID=105954652
|
|
|
|
|
|
|
|
|
| |
file rather than a ParserInputSource.
This is part of a series of changes with the net result being that we open, read, and parse each BUILD file exactly once.
--
MOS_MIGRATED_REVID=105911557
|
|
|
|
|
|
|
| |
FileStatus#isSpecialFile to help disambiguate between a regular file and a special file, since the file size of a special file cannot be trusted.
--
MOS_MIGRATED_REVID=105903622
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105851371
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105844221
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*** Reason for rollback ***
Uses tests that don't run on Bazel
*** Original change description ***
Implement aspect(...) Skylark function.
--
MOS_MIGRATED_REVID=105808850
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
While a normal set is theoretically sufficient, it can cause hard-to-reproduce
problems. In particular, the iteration order of the expansion result depends
on the iteration order of the requested targets. If there are multiple
requests for the same set of targets, but with different orders, the results
would depend on which request was made first. If a downstream function also
inadvertantly depends on the iteration order, it can be hard to debug why it
ended up with a different order than it requested.
Alternatively, we could sort the result before returning it. We generally
expect the key set to be smaller than the result set, so this should be
ever so slightly more efficient.
--
MOS_MIGRATED_REVID=105765514
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105742091
|
|
|
|
|
|
|
|
| |
This is still not hooked up, but it's now passing all the relevant tests that
apply directly to the LoadingPhaseRunner; there are still a couple of TODOs.
--
MOS_MIGRATED_REVID=105739507
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105734972
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105648425
|
|
|
|
|
|
|
| |
Minimize indirection wrt singletons, turns out we had one that was completely unused.
--
MOS_MIGRATED_REVID=105621494
|
|
|
|
|
| |
--
MOS_MIGRATED_REVID=105596479
|
|
|
|
|
|
|
|
|
| |
since root causes are associated with top-level targets and labels, and the filter is for all top-level targets and labels.
I noticed this when a huge --noanalyze build spent most of its time in filtering here. The passed-in "collection" was a list, which meant that we could have sped it up by using a set, but why not just get rid of it all.
--
MOS_MIGRATED_REVID=105536485
|
|
|
|
|
|
|
|
|
| |
former uses the latter). This way we parse BUILD files exactly once.
This is part of a series of changes with the net result being that we open, read, and parse each BUILD file exactly once.
--
MOS_MIGRATED_REVID=105528075
|