Concurrent File Operation Race Condition Verification Benchmark

#race condition detection #deadlock verification #concurrency bug fixing #concurrent file operation safety #file system reliability #concurrency defect detection

5000
0.4GB
JSON
CATL
MOBIUSI INC

Updated:2026-11-05

AI Analysis & Value Prop

This benchmark contains source code tasks and repair objectives involving concurrent reads and writes, lock management, cache updates, file closure, and metadata access in distributed file systems. Its task contexts and test setups are designed to assess whether code exhibits races, deadlocks, or inconsistent shared state under concurrent execution. It supports evaluation of code agents' defect detection, repair, and execution verification capabilities.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
source_code	string	Concurrent file operation source code or code snippets to be analyzed and verified.
task_context	string	Describes the file system scenario, concurrent participants, thread or process relationships, and relevant runtime conditions.
repair_objective	string	Describes the concurrency defect to fix, the intended direction of changes, and the correctness requirements.
concurrency_risks	array	A list of concurrency risk types identified from the source code and task description.
validation_criteria	array	Observable conditions and test outcomes used to determine whether a repair is effective.
programming_language	string	The primary programming language identified from the source code.
concurrency_test_setup	string	Describes the test triggers, concurrent scheduling, read and write order, lock operations, cache state, or file lifecycle conditions used for verification.
deadlock_risk_detected	boolean	Indicates whether lock ordering, release paths, and blocking relationships suggest a deadlock risk.
race_condition_detected	boolean	Indicates whether analysis or testing found unsynchronized concurrent reads, writes, or state updates.
expected_correct_behavior	string	Describes the behavior that should remain correct under concurrent execution, exceptional paths, and file lifecycle changes after repair.
operation_and_lock_summary	string	Summarizes key interactions among file reads and writes, lock acquisition and release, cache updates, file closure, and metadata access.
repair_verification_summary	string	Summarizes repair performance in concurrency tests and notes boundary conditions or residual risks that may require attention.
state_consistency_risk_detected	boolean	Indicates whether concurrent operations may leave caches, file metadata, file handles, or other shared state inconsistent.

Compliance Statement

Authorization Type	Proprietary - Commercial AI Training License (No Redistribution)
Commercial Use	Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization	No PII, no real company names, simulated scenarios follow industry standards
Compliance System	Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs

Frequently Asked Questions

Which concurrent file operation scenarios are covered?: Tasks focus on concurrent reads and writes, lock acquisition and release, cache updates, file closure, and metadata access in distributed file systems. Context and test setup details help analyze how these operations interact under concurrency.

Which code verification capabilities can it assess?: The benchmark can assess detection of data races, deadlock risks, and inconsistent shared state. It can also test whether a code agent proposes repairs that meet the expected correctness requirements and whether those repairs pass the stated validation criteria.

What information is included in the dataset fields?: Fields cover source code, task context, repair objectives, concurrency test setup, programming language, concurrency risk types, operation and lock summaries, expected behavior, validation criteria, and repair verification summaries. The information available may vary by record.

How can I obtain the data, and what are the licensing terms?: Check the dataset product page or the provider's designated channel for access instructions, delivery format, and licensing terms. Before use, confirm whether the license permits training, evaluation, or redistribution of the code examples; this description does not replace the formal license.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2026,
  title={Concurrent File Operation Race Condition Verification Benchmark},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/d21678c9aff545da6525f83edcfccbf6},
  urldate={2026-10-14},
  keywords={concurrent file operations,race condition detection,distributed file systems,concurrent code verification,deadlock risk analysis,file lock management,cache consistency,code agent evaluation},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task