[libc++] Use native wait in std::barrier instead of sleep loop #171041
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
huixie90
force-pushed
the
hxie/barrier_native
branch
from
6ce071d to
cd942d4
Compare
huixie90
force-pushed
the
hxie/barrier_native
branch
from
cd942d4 to
19eaea4
Compare
llvmbot
added
the
libc++
Member
|
@llvm/pr-subscribers-libcxx Author: Hui (huixie90) Changesnot sure why but the current Full diff: https://github.com/llvm/llvm-project/pull/171041.diff 1 Files Affected:
diff --git a/libcxx/include/barrier b/libcxx/include/barrier
index 5f9b471f01741..10d40f32919c8 100644
--- a/libcxx/include/barrier
+++ b/libcxx/include/barrier
@@ -142,8 +142,7 @@ public:
return __old_phase;
}
_LIBCPP_HIDE_FROM_ABI void wait(arrival_token&& __old_phase) const {
- auto const __test_fn = [this, __old_phase]() -> bool { return __phase_.load(memory_order_acquire) != __old_phase; };
- std::__libcpp_thread_poll_with_backoff(__test_fn, __libcpp_timed_backoff_policy());
+ __phase_.wait(__old_phase, std::memory_order_acquire);
}
_LIBCPP_HIDE_FROM_ABI void arrive_and_drop() {
__expected_adjustment_.fetch_sub(1, memory_order_relaxed);
|
ldionne
changed the title
ldionne
approved these changes
Dec 12, 2025
Member
ldionne
left a comment
ldionne
left a comment
There was a problem hiding this comment.
LGTM, thanks! This seems like an obvious improvement.
Collaborator
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/169/builds/18010 Here is the relevant piece of the build log for the reference |
anonymouspc
pushed a commit
to anonymouspc/llvm
that referenced
this pull request
Dec 15, 2025
…171041) For some reason, the current `std::barrier`'s wait implementation polls the underlying atomic in a loop with sleeps instead of using the native wait. This change should also indirectly fix the performance issue of `std::barrier` described in llvm#123855. Fixes llvm#123855
mahesh-attarde
pushed a commit
to mahesh-attarde/llvm-project
that referenced
this pull request
Dec 19, 2025
…171041) For some reason, the current `std::barrier`'s wait implementation polls the underlying atomic in a loop with sleeps instead of using the native wait. This change should also indirectly fix the performance issue of `std::barrier` described in llvm#123855. Fixes llvm#123855
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
For some reason, the current
std::barrier's wait implementation polls the underlying atomic in a loop with sleeps instead of using the native wait.Before this change, on the CI runner , lost_wakeup.pass.cpp was the slowest test , which uses
std::barrier===
macOS CXX23
Before:
After
===
Linux CXX23:
Before
After:
No longer in the slowest tests list
===
Windows:
Before
After
No longer in the slowest test list
This change should also indirectly fix the performance issue of
std::barrierdescribed in #123855.Fixes #123855