summaryrefslogtreecommitdiff
path: root/test/rollback_bench.zig
Commit message (Collapse)AuthorAge
* replica clients over a socket: the session streams each its dump and every ↵Gabriel Schneider2026-10-30
| | | | | | | | | | | | unit after it (replica.zig `Peers`, a writer and a reader thread a client, units at each step boundary, a client's input stepped there in its own view), and `pardes --attach=<socket>:<name>` and `--attach=<host>:<name>` (through `ssh -tt -L`) are a client in the tty host, its keys applied at once by its own replica and drawn by its own process wire.zig version 15: welcome (the client's id and view, the dump's length, the session's UTC offset and digest), dump pieces, units (with the authority's clocks), terminal messages both ways, a check. A unit carries one digest, at its end (`unit_digests`). The client: tty.zig `Shell.remote`, `feed` (a terminal's input to the replica and the session); the session's clock guessed from the last unit's (tape.zig `Guess`); a pump woken only by a client's post reads no new time (`mail_only`), and a unit of another's ending where the last ended leaves the speculation as it is (`neutral`); the rollback copy taken while idle (`Replica.ready`); views a frontend here does not draw left undrawn. A lagging client never holds the session: its queue is cut off past 64 MiB (PARDES_REPLICA_LAG_KB). Detach in a client ends that client; a Restore lets every client go. daemon.zig: a replica the daemon hands over is the session's loop's (a note), not a stub thread's. A window session refuses one in words. test/daemon.py `--clients`: two clients (one through the delay relay) type into a detached session and agree with it, one killed outright, one detached, a third typing into a shell. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* phase 5 part A: the speculation and rollback engine, in one process: a ↵Gabriel Schneider2026-10-29
| | | | | | | | | | | | | | | | walker copy put back where it was, replicas that apply their own input at once and take the authority's tape as the truth, and a harness of typists over delayed links src/rollback.zig: the replica core's allocator (`Epoch`) knows every allocation, and while a copy is held puts off frees of what was there before and refuses to grow it in place, so what the core held stays where it was. The copy (`Copy`) is the digest's walk saving the bytes of every allocation it reaches, field by field (the derived caches and this process's own too: a safe build's free writes over what it frees, and the frame arena points at what the epoch then frees), but the queues a step drains and the frame statistics; a file's text, its line index and its undo snapshots are pinned instead (made anew at each edit, never written; now freed raw everywhere), checked unchanged in a safe build. Put back, each allocation's bytes go back over it, the epoch's allocations are freed and its put-off frees forgotten; let go, they are made. Tree-sitter's trees, the C heap's, are held the same way (syntax.held_trees). src/replica.zig: an `Authority` records its core's every input to a tape in memory and hands each client's input back as a unit (origin, sequence number, its tape); a `Replica` applies its own input at once, as the authority will (update, then pump), recorded by its own tape (tape.zig's speculation mode), and for each unit in order drops its pending head when the authority recorded the same (inputs, host calls and answers: the fast path), else puts the copy back, replays the units since and this one (every digest checked) and applies its pending inputs again. A host call or a syscall stops a speculation: it is answered harmlessly, undone when it returns, and the input waits for the authority's answer; a clock or environment read is a guess the comparison checks; an emulator is the authority's (terminal.zig's writers stop a speculation; a safe build asserts none changed). The view's events (a resize, the pointer) wait. Messages are bytes: no transport yet. Each client has a view of its own (src/views.zig): the authority goes into it before the client's input (`toView`, taped), the replica's speculation the same, and the copy walks the views not live too (`parked`, each pane's `pv`). test/replicas.zig (`zig build replicas`): one authority and N replicas, a view each (`--shared` puts them in one), each client typing monkey.zig's mix at 50-200 ms over its own link (half the RTT plus 10% jitter, in order), the authority's pty echoing 2 ms later; gates: no replica diverges, every replica's digest is the authority's at quiescence; reported per RTT: inputs applied at once, stopped, waiting, fast paths, rollbacks (mispredicted, foreign), and what a speculative apply, a copy and a rollback cost. 3 clients x 100 inputs at 30, 80 and 207 ms: no divergence over 12 Debug seeds and 20 ReleaseFast ones with a filled heap; ReleaseFast p99 a speculative apply under 0.2 ms, a rollback under 1.6 ms. `zig build rollback-bench` times the real copy now (take, release, an edit, restore). What the replicas found, fixed: a release build reads what the heap had where Debug's fill made two cores agree: `col_panes` and `restore_buf` digested undefined (zeroed now), `terminal.State.reply_dropped` never set at create, `Text.sels` and `PendingPipe.ranges` digested past their length (a walker `len` cuts a slice too). PARDES_REPLICA_FILL fills a replica's allocations to show what is left. Co-Authored-By: Claude Opus 5.5 <[email protected]>
* rollback-bench and fork_stall.py: what a replica's rollback copy is (a walk ↵Gabriel Schneider2026-10-29
of ~100 KB once file texts are pinned) and what a fork stalls the authority for, which settle open points 3 and 5 `zig build rollback-bench -Dplatform=tty -Doptimize=ReleaseFast` (test/rollback_bench.zig) builds a typical session (a shell with 1 MB of scrollback, three 60 KB files and a 3 MB one, 40 edits) and a big one (14 panes, 18 MB, 4x3 MB and 8 small files, 8 MB of scrollback, 200 edits) in process, then measures the state the digest walk visits, the part of it that is a file's text and its undo snapshots (immutable once made: an edit puts a new buffer in place, a snapshot is a copy, so a copy pins them), and the walk again with those taken out, plus a memcpy of what is left. Typical: 36.3 MB walked, 80 KB to copy, about 0.02 ms; big: 282.5 MB, 96 KB, about 0.045 ms. Then a fork of the big session (431 MB RSS: 3.4 to 8.3 ms) and 64 MB written by the parent under the child (~28 ms of copy-on-write). test/fork_stall.py does the fork alone, on any system with a Python (the FreeBSD box): N MB written, then fork() timed and 64 MB written under the child, with Linux's THP as set and with 4 KiB pages. 1 GB: 13.3 to 19.6 ms here, 20.4 ms on FreeBSD 15.1; the copy-on-write after it ~1.3 us a page on both. Co-Authored-By: Claude Opus 5.5 <[email protected]>