MattJackson 6abc25a7e8 iter7: WRITEBACK_CHUNK_BYTES 32 -> 64 MiB ...

iter6 (8 MiB chunks) regressed mean -8.2 MB/s vs iter4 (32 MiB).
Trend: bigger = better in this workload, against the page-age
theory. Try 64 MiB to see if fewer/larger syncs raise mean further.
0.21.14 went to 128 MiB and was reverted; 64 is between.

2026-05-17 08:25:44 -07:00

..

file_sector_source

Revert "iter2: restore FileSectorSource readahead buffer (32 MiB)"

2026-05-17 07:38:34 -07:00

sink

io+mux: phase 3 — streaming sinks + sequential container muxers

2026-05-13 20:17:33 -07:00

writeback

Revert "io/writeback: medium-agnostic by construction, no DONTNEED, single detection point"

2026-05-16 17:02:20 -07:00

writeback_file

iter7: WRITEBACK_CHUNK_BYTES 32 -> 64 MiB

2026-05-17 08:25:44 -07:00

bounded.rs

v0.20.6: io::bounded — halt-safe wrapper for blocking syscalls

2026-05-13 14:22:24 -07:00

byte_channel.rs

io: phase 1 buffering — read-side flatness

2026-05-13 19:48:23 -07:00

mod.rs

io+platform: phase 2 — sink trait split + fs_type detection

2026-05-13 19:58:54 -07:00

pipeline.rs

iter6: revert depth bump + WRITEBACK_CHUNK_BYTES 32->8 MiB

2026-05-17 08:09:57 -07:00

writeback.rs

v0.17.10: bounded-cache writeback pipeline for big sequential writes

2026-05-08 19:54:25 -07:00