Fix #2563: Unconditionally emit mixin forwarders #6081
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
smarter
force-pushed
the
mixin-force-forwarders-2
branch
from
ab689a4 to
2b27e3e
Compare
|
test performance please |
|
performance test scheduled: 1 job(s) in queue, 1 running. |
|
Performance test finished successfully: Visit http://dotty-bench.epfl.ch/6081/ to see the changes. Benchmarks is based on merging with master (3c9935f) |
odersky
approved these changes
Mar 17, 2019
smarter
force-pushed
the
mixin-force-forwarders-2
branch
from
2b27e3e to
e70b6e4
Compare
This controls when mixin forwarders are generated like the homonymous flag in scalac, for now we default to "junit" to keep our existing behavior, next commit will change it to "true" to match scalac.
This is what Scala 2.12+ does for cold performance reasons (see scala#5928 for more details) and we should align ourselves with them when possible. About two years ago in scala#2563, Dmitry objected that we may not need to do that if we found another way to get performance back, or if newer JDKs improved the performance of default method resolution. It doesn't look like these things have happened so far (but there's some recent glimmer of hope here: https://bugs.openjdk.java.net/browse/JDK-8036580). Dmitry also said "I don't recall triggering bugs by emitting more forwarders rather then less.", but in fact since then I've found one case where the standard library failed to compile with extra forwarders causing name clashes, requiring a non-binary-compatible change: scala/scala@e3ef657. As of scala#6079 this is no longer a problem since we now emit mixin forwarders after erasure like scalac, but it still seems prudent to emit as many forwarders as scalac to catch potential name clash issues.
Now that -Xmixin-force-forwarders defaults to true, tests that showcase some behavior of mixin forwarders can be simplified since we need less indirections to generate them.
smarter
force-pushed
the
mixin-force-forwarders-2
branch
from
e70b6e4 to
22db0c0
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR is a draft because it depends on #6079, only the last three commits are new.