1.1.0 release train - #3307
Draft
pivovarit wants to merge 28 commits into
Draft
Conversation
While the comment is correct, it's way more pragmatic to explicitly mark the field as _volatile_ (cherry picked from commit 6d3be22)
The old implementation independently called replace on both front and rear, causing up to two replacements instead of one. (cherry picked from commit e50089d)
Queue#flatMap produced incorrectly ordered results when the queue had elements in its internal rear list (cherry picked from commit 1b22db5)
(cherry picked from commit 8bd9c6b)
(cherry picked from commit 2b77936)
(cherry picked from commit 710325c)
(cherry picked from commit 8305062)
(cherry picked from commit d52a062)
(cherry picked from commit 2e5f3c3)
(cherry picked from commit b9dfea1)
(cherry picked from commit 45f8698)
- stop cancellation from dispatching queued or newly registered completion callbacks - make `andThen` use a cancellable result future so cancelling the chained future prevents the delayed side-effect action - add regression coverage for direct callbacks and cancelled `andThen` chains Fixes #2552 - `JAVA_HOME=/opt/homebrew/Cellar/openjdk/25.0.2/libexec/openjdk.jdk/Contents/Home PATH=/opt/homebrew/Cellar/openjdk/25.0.2/libexec/openjdk.jdk/Contents/Home/bin:$PATH ./mvnw -pl vavr -DskipCheckstyle -DskipCoverage -Dmaven.repo.local=$PWD/.m2/repository -Dtest=io.vavr.concurrent.FutureTest test` - `git diff --check` --------- Co-authored-by: Puneet Dixit <236133619+puneetdixit200@users.noreply.github.com> Co-authored-by: Deepak kudi <deepakkudi23@adsl-172-10-9-116.dsl.sndg02.sbcglobal.net> (cherry picked from commit dffe169)
(cherry picked from commit a7166cd)
(cherry picked from commit 31eff71)
(cherry picked from commit 1d1f3a2)
before:
```
distinct=66000 total=900000 -> size=66000 in 178172 ms
```
after:
```
distinct=66000 total=900000 -> size=66000 in 232 ms
```
reproducer:
```
static void shakespeareStyleWorkload() {
int distinct = 66_000, total = 900_000;
int[] corpus = new int[total];
for (int i = 0; i < total; i++) {
corpus[i] = i % distinct;
}
long t0 = System.nanoTime();
LinkedHashMap<Integer, Integer> m = LinkedHashMap.empty();
for (int word : corpus) {
m = m.put(word, m.get(word).getOrElse(0) + 1);
}
long t1 = System.nanoTime();
System.out.printf("distinct=%d total=%d -> size=%d in %.0f ms%n",
distinct, total, m.size(), (t1 - t0) / 1_000_000.0);
}
```
Queue and Vector make very different tradeoffs, and Queue's tradeoffs are wrong for this job.
`LinkedHashMap` doesn't use it like a queue. It uses it as an insertion-ordered sequence it reads many times without consuming. That breaks the amortization.
Vector keeps everything already in the correct order. Nothing is stored backwards, so nothing ever has to be flipped before reading. Getting the first element is instant. It uses slightly more memory, but it sounds like a reasonable trade-off.
Before:
```
size per-call (us) ratio vs prev
20000 95,305 -
40000 186,741 1,96x
80000 295,748 1,58x
160000 1153,626 3,90x
320000 1878,587 1,63x
```
After:
```
size per-call (us) ratio vs prev
20000 0,278 -
40000 0,299 1,07x
80000 0,134 0,45x
160000 0,108 0,80x
320000 0,122 1,13x
```
----
reproducer:
```
private static double measureFirstElementNanos(LinkedHashSet<Integer> set, int reps) {
int sink = 0;
long t0 = System.nanoTime();
for (int r = 0; r < reps; r++) {
sink ^= set.iterator().next();
}
long t1 = System.nanoTime();
if (sink == 0xDEADBEEF) {
System.out.print("");
}
return (t1 - t0) / (double) reps;
}
```
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
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.
No description provided.