Landing: 6f32166346de

← All landings

Project / Subsystem

gcc / range-op-float

Date

2026-07-31

Author

Jakub Jelinek

Commit

6f32166346de2d1c0f6253e4fe37e8a6edac655d

Source

github

Perf win

Breaking

View on GitHub →

All attributes

project: gcc
subsystem: range-op-float
patch_id: —
commit_hash: 6f32166346de2d1c0f6253e4fe37e8a6edac655d
source_type: github
headline: range-op-float: Improve infinity handling in reverse floating-point operations
tldr: Fixes incorrect range calculations for reverse floating-point operations involving infinity.
author: Jakub Jelinek
outcome: committed
performance_win: false
breaking_change: false
series_id: —
series_parts: []
tags: • compiler
• optimization
• floating-point
• range analysis
discussion_id_link: —
bugzilla_pr: —
url: https://github.com/gcc-mirror/gcc/commit/6f32166346de2d1c0f6253e4fe37e8a6edac655d
date: 2026-07-31T00:00:00.000Z

This commit improves the handling of infinity in reverse floating-point operations, addressing miscompilations that occurred with specific range analyses. Previously, ranges like [-inf, -inf] were incorrectly propagated in reverse operations, leading to wrong results when converting between floating-point types. The fix now correctly widens the bounds of infinity ranges by 1 unit in the penultimate representation, ensuring accurate results for conversions and other operations.