[Progress Report] libgomp Optimizations for Scheduler Guided OpenMP Execution in Cloud VMs

Project / Subsystem

gcc / gcc

Date

2026-08-27

Proposer

Roy Fru <royfru44@gmail.com>

Source type

public_inbox

Consensus

Proposed

Sentiment

—/10

Technical tradeoffs

  • • Replacing read/pread with mmap reduces system call overhead but requires careful handling of memory mapping and unmapping.
  • • Transitioning from uprobes to USDT provides a more stable tracing interface but may require changes in how eBPF programs are written and attached.
  • • Shared metadata and dynamic IVSHMEM allocation add complexity to the runtime but could improve scalability and resource utilization.

All attributes

project
gcc
subsystem
gcc
patch_id
—
discussion_id
CAAk=M9cNW__YSJUUL4yiG=jW5toR53J3w0TizehtWmmJh=kcfg@mail.gmail.com
source_type
public_inbox
title
[Progress Report] libgomp Optimizations for Scheduler Guided OpenMP Execution in Cloud VMs
headline
Libgomp optimizations for scheduler-guided OpenMP execution
tldr
Roy Fru reports progress on optimizing libgomp for cloud VMs by replacing read operations with mmap, adopting USDT probes, and testing with NPB benchmarks.
proposer
Roy Fru <royfru44@gmail.com>
consensus
Proposed
outcome
proposed
sentiment_score
—
technical_tradeoffs
  • • Replacing read/pread with mmap reduces system call overhead but requires careful handling of memory mapping and unmapping.
  • • Transitioning from uprobes to USDT provides a more stable tracing interface but may require changes in how eBPF programs are written and attached.
  • • Shared metadata and dynamic IVSHMEM allocation add complexity to the runtime but could improve scalability and resource utilization.
series_id
—
series_role
standalone
series_parts
[]
tags
  • • optimization
  • • OpenMP
  • • libgomp
  • • virtualization
  • • tracing
bugzilla_url
—
date
2026-08-27T00:00:00.000Z

[Progress Report] libgomp Optimizations for Scheduler Guided OpenMP Execution in Cloud VMs

Roy Fru provides a progress report on optimizing libgomp for OpenMP execution in cloud VMs. Key advancements include replacing read operations with mmap for IVSHMEM pages to avoid system calls, and transitioning from uprobe-based to USDT probes for more stable tracing. The integration has been tested with NPB benchmarks, and future work may involve a shared metadata section for VMs.