Supporting non-power of 2 zoned devices
summary
Adds support for zoned devices whose usable blocks per zone are not powers of two by introducing a device‑mapper target that emulates power-of-two zones so existing filesystems keep working. It detects and hides LBA gaps by splitting bios that cross the usable capacity, returning zeros for emulated ranges, and forbidding writes there. Patchset is v13, tested with blktests and zonafest, shows no hot‑path regression; native block/NDME support lands first, filesystems get phased native support later.
AI-written summary · about
chapters (11)
- 0:00Introduction and agenda
- 1:02Definitions and background
- 3:30Zone capacity and LBA gaps
- 6:01Work on non-power of two zones
- 14:29Motivation for change
- 16:14Historical attempts and goals
- 18:21Current approach and phases
- 19:22Performance regression testing
- 20:46Device mapper target dm-zoned
- 22:54Future work and acknowledgments
- 24:44Q&A session
related talks
25mUsing Zoned Block Devices in Fedora
31mFrom Open-Channel SSDs to Zoned Namespaces
39mThe Power of the Device Mapper - From Dm-cache To Dm-zoned
14mUnleashing Zoned UFS: Cross-Layer Optimizations for Next-Generation Mobile Storage
21mAdvanced block storage test devices
14mZNS+: Advanced Zoned Namespace Interface for Supporting In-Storage Zone Compaction
32mUsing Linux Block Integrity in Building and Testing Storage Systems
45mData Placement in Linux: Evolving Block and File I/O