Implementing Case-insensitive File Systems in the Linux Kernel
Talkat Open Source Summit Japan + AI_dev: Open Source GenAI & ML Summit + Automotive Linux Summit 2025 · 2025-12-08..12-10·
summary
Linux now supports Unicode-aware case folding inside the kernel so filesystems can offer true case-insensitive behavior; ext4 added full Unicode CI in 2018 and f2fs, bcachefs, tmpfs and overlayfs have followed. The kernel provides a casefold/normalization library and uses hashed lookups for performance, avoiding slow user-space scanning, but CI is opt-in per-directory, allowed only on empty dirs, and overlayfs had to store names twice to merge layers. Complexities include NFD normalization, full vs simple folding, locale anomalies like Turkish I, and past regressions that made files unreachable.
AI-written summary · about
chapters (10)
- 0:00Case-insensitive file systems intro
- 8:34Why Linux is case-sensitive
- 9:37Need for case-insensitive FS
- 14:13Unicode case folding
- 21:41Normalization rules
- 23:28Kernel Unicode support
- 24:10Implementing case-insensitive dirs
- 26:57Limitations of case-insensitive FS
- 31:10Summary
- 31:35Q&A
related talks
28mExt4: A Case-insensitive Linux Filesystem
17mCoFS: A Filesystem for Fast Container Startup
36mFlash Friendly File System
19mExtension Framework for File Systems in User space
26mImprove FUSE Filesystem Performance and Reliability
33mEncryption for filesystems with advanced features new fscrypt functionality
14mAdvancing Data Integrity in Linux
45mData Placement in Linux: Evolving Block and File I/O