Parallelize the exhaustive float32 sweeps across hardware threads (~75-88x)#383
Merged
Merged
Conversation
Same std::thread split as exhaustive32_midpoint; preserves each test's existing failure behavior (abort for exhaustive32, stop-flag for exhaustive32_64).
fcostaoliveira
changed the title
Contributor
Author
|
Extended to the sibling sweeps as offered — 96-core timings (gcc), all pass and clean under -Werror on gcc + clang:
Happy to drop the two extra files back out if you'd prefer to keep this PR to just the midpoint test. |
Member
|
It is fine. Merging. |
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
exhaustive32_midpointsweeps all 2³² float bit-patterns one at a time, which takes~30 min on a fast machine. The values are independent, so this splits the range across
std::thread::hardware_concurrency()threads, with an atomic flag for fail-fast. Theper-value work is factored into a
check_word()helper; the checks and pass/failbehavior are unchanged.
On a 96-core machine the runtime drops from ~1900 s to ~25 s (≈75×), identical result:
Notes:
FASTFLOAT_EXHAUSTIVE(off by default and notbuilt in CI), so this has no CI impact — it's purely a faster local/dev sweep.
tests/CMakeLists.txtnow linksThreads::Threads(withTHREADS_PREFER_PTHREAD_FLAG).-Werror -Wall -Wextra -Weffc++ -Wconversion -Wsign-conversion -Wshadowset on gcc and clang; clang-format clean.The same pattern applies to the sibling sweeps (
exhaustive32,exhaustive32_64); Ikept this PR to one test for review, happy to extend it if you'd like.