Release v2.21.0

- Protocol 2.21.0: STATUS_ERROR_DETAIL rejection reasons and server-contacting --dry-run
- Daemon per-module/per-host caps and cross-process auth lockout
- Config X-macro serialization, authorized_root single-owner, Data charge ownership, receiver pipeline move
- Security audit hardening (SSH injection, FIFO/inplace, zstd DoS, TLS, dry-run oracle, bounds)
- Pre-auth basis_count NULL-deref fix; benchmark and nix-shell improvements
- Tested: unit, integration, ASan/UBSan, valgrind, fuzz, coverage (CI green)

CHANGELOG.md

@@ -4,41 +4,7 @@ All notable changes to FastSync are documented here. Versions match
 `PROTOCOL_VERSION` (printed by `fastsync --version`); the client and server must
 run the same version because the handshake is strict.
 
-## [Unreleased]
-
-### Added
-
-- **Server-contacting `--dry-run` (protocol 2.21.0).** `--dry-run` now performs
-  a real handshake with a remote/daemon receiver and reports exactly what WOULD
-  change based on receiver state (existing destination files, mtimes, checksums,
-  basis dirs). The wire config carries the dry-run intent (`Config.dry_run`) and
-  the receiver answers each per-file check with `STATUS_DRY_RUN_TRANSFER` (would
-  transfer) or `STATUS_OK` (already up to date); the sender prints the
-  would-transfer set and its trailer without sending any file data. The receiver
-  performs the normal read-only incremental decision but mutates nothing: no temp
-  files, writes, renames, deletes, metadata/xattr/chown, or directory creation.
-  A plain local destination (no explicit `--server-port`/remote) keeps the
-  original client-side dry-run. Would-delete reporting for `--delete*` is
-  deferred to a follow-up; dry-run never deletes.
-
-### Security
-
-- Enforce the daemon's per-module `max connections` cap and add a global
-  `max connections per host` cap plus a cross-process `auth lockout`
-  (`auth lockout threshold` / `auth lockout duration`). Because the listener
-  forks one child per connection, the counters live in an anonymous shared
-  mapping created before the accept loop and reclaimed by the parent's
-  `SIGCHLD` handler, so the per-module, per-source and auth-failure state is
-  shared across every child (including after `SIGKILL`). The per-source table
-  now has a bounded lifetime (expired-lockout/idle entries are reclaimed, with a
-  rate-limited warning when it is genuinely full), and the occupancy counters are
-  re-derived from the shared slot table on every child exit so a child killed
-  mid-registration cannot leak a count. Trusted loopback peers are exempt from the
-  per-host cap and the auth lockout (they share one address); clients behind a
-  shared NAT/proxy still share a single per-host budget and lockout, which is
-  documented.
-
-## [2.21.0] - 2026-09-13
+## [2.21.0] - 2026-09-14
 
 ### Added
 
@@ -51,12 +17,97 @@ run the same version because the handshake is strict.
   the reason into a thread-local buffer exposed by `protocol_last_error()`. The
   detail body is always consumed, so the stream cannot desynchronize, and
   messages are sliced to `MAX_ERROR_DETAIL_BYTES` (4096) on send.
+- **Server-contacting `--dry-run` (protocol 2.21.0).** `--dry-run` now performs
+  a real handshake with a remote/daemon receiver and reports exactly what WOULD
+  change based on receiver state (existing destination files, mtimes, checksums,
+  basis dirs). The wire config carries the dry-run intent (`Config.dry_run`) and
+  the receiver answers each per-file check with `STATUS_DRY_RUN_TRANSFER` (would
+  transfer) or `STATUS_OK` (already up to date); the sender prints the
+  would-transfer set and its trailer without sending any file data. The receiver
+  performs the normal read-only incremental decision but mutates nothing: no temp
+  files, writes, renames, deletes, metadata/xattr/chown, or directory creation.
+  A plain local destination (no explicit `--server-port`/remote) keeps the
+  original client-side dry-run. Would-delete reporting for `--delete*` is
+  deferred to a follow-up; dry-run never deletes.
+- Daemon `max connections per host` (per-source-IP concurrent cap, default 0 =
+  unlimited), `auth lockout threshold` (default 10; 0 disables) and
+  `auth lockout duration` (default 300 s) config keys.
+- `fastsync-server --allow-super` opt-in for a privileged standalone TCP server;
+  without it a root standalone receiver forces super-user activities off (device
+  nodes, `--write-devices`, ownership). The `--stdio` SSH argv is client-composed,
+  so super activities always stay off there.
 
 ### Changed
 
+- Config wire fields are now declared once in an X-macro table
+  (`CONFIG_WIRE_FIELDS` in `src/shared/config.h`) that generates the struct
+  members, defaults, and the send/receive sequence, removing the manual
+  six-site field sync. Wire bytes and `PROTOCOL_VERSION` are unchanged.
 - `receive_incremental_check()` (the per-file `STATUS_CHECK` fast path) is split
   into small static helpers with a short linear orchestrator. Pure refactor: the
   wire byte stream and all cleanup are unchanged.
+- `authorized_root` state has a single owner (`utils.c`) with read accessors; the
+  duplicated statics in `file.c` and the server were removed.
+- `Data` records its owning `ProtocolSession` so its memory charge is returned to
+  the session that reserved it, regardless of the destroying thread.
+- The receiver pipeline moved out of `shared` into `server/receiver_pipeline.[ch]`;
+  the build now uses explicit `fastsync_shared` / `fastsync_client_core` /
+  `fastsync_server_core` targets instead of a GLOB, and the client no longer links
+  server code.
+- The benchmark tool generates the requested random/compressible data mix
+  accurately, verifies each transfer before recording it, computes correct
+  percentiles, adds a MB/s column, handles `tc`/netem without requiring `sudo`
+  when already root, builds into a dedicated `build-bench/` directory, and adds a
+  `--warm` incremental-transfer mode.
+- The `nix-shell` dev environment provides the full toolchain (clang-format,
+  cppcheck, pytest-xdist, OpenSSH, rsync, iproute2, valgrind, lcov) and no longer
+  builds on entry.
+
+### Security
+
+- Enforce the daemon's per-module `max connections` cap (0 = unlimited) and add
+  the shared per-source `max connections per host` cap plus a cross-process
+  `auth lockout`. Because the listener forks one child per connection, the
+  counters live in an anonymous shared mapping created before the accept loop and
+  reclaimed by the parent's `SIGCHLD` handler, so the per-module, per-source and
+  auth-failure state is shared across every child (including after `SIGKILL`). The
+  per-source table has a bounded lifetime (expired/idle entries are reclaimed,
+  with a rate-limited warning when genuinely full), and the occupancy counters are
+  re-derived from the shared slot table on every child exit. Trusted loopback
+  peers are exempt (they share one address); clients behind a shared NAT/proxy
+  share a single per-host budget and lockout, which is documented.
+- Hardening from a full security audit:
+  - Fail a truncated zstd frame instead of spinning forever (remote DoS).
+  - Open receiver destination/basis/hard-link entries `O_NONBLOCK` so a
+    client-planted FIFO cannot block a worker indefinitely.
+  - Require a regular file before `--inplace` writes, closing a FIFO-hang and a
+    raw-device write that bypassed the `--write-devices` gate.
+  - Reject SSH destinations whose user/host begins with `-` and insert `--` before
+    the host token, closing `-o ProxyCommand=…` argument injection (RCE).
+  - Gate client `--force` recursive removal behind the server `--allow-delete`
+    policy.
+  - Reject empty `hosts allow`/`hosts deny`/`auth users` values instead of
+    silently meaning "unrestricted".
+  - Restrict TLS 1.2 to AEAD suites and set server cipher preference; load the
+    private key TOCTOU-safely from an `O_NOFOLLOW` fd; verify IP literals against
+    IP SANs; guard client-cert CN truncation.
+  - Make `--dry-run` content-blind: it neither reads destination files nor
+    hashes basis files, removing a 1-bit content oracle against `read only`
+    modules.
+  - Bound glob matching (iterative DP, no exponential backtracking) and bound
+    line reads for filter/`--files-from`/pattern files.
+  - Gate `system.posix_acl_*` xattrs on `--acls` and charge decompression/chunk
+    allocations against the per-connection memory budget.
+
+### Fixed
+
+- Pre-auth NULL dereference in `config_delete()` when an over-long
+  `basis_count` (and the analogous count fields) was received and then failed
+  validation; received counts are now validated before being published.
+- Leaked inherited `Data` in the forked compression-truncation unit test
+  (valgrind definite leak).
+- `receive_status()` no longer loses a captured rejection reason when owed
+  keepalives are drained.
 
 ## [2.20.0] - 2026-09-13
 

README.md

@@ -178,6 +178,7 @@ transfer is never aborted.
 | `--ca <path>` | TLS CA certificate file for verification (PEM) |
 | `--destination-root <path>` | Authorized destination root (default: `.`) |
 | `--allow-delete` | Permit manifest deletion |
+| `--allow-super` | Standalone TCP listener only: keep super-user activities enabled for a **root** receiver. Without it a root standalone server forces `SUPER_MODE_OFF`, so client `--devices`/`--write-devices`/`--super` and client-chosen ownership requests are skipped/refused. **Rejected with `--stdio`** (the SSH remote argv is client-composed, so a client could otherwise pass it and defeat the secure default; operators exposing `fastsync-server --stdio` over SSH must use a forced command if the default must hold). No effect when not root. |
 | `--allow-unauthenticated` | Permit plaintext TCP clients. For an `auth users` module this opts in **loopback plaintext only**; remote auth still requires verified TLS, so the flag never permits remote plaintext auth. |
 | `-v, --verbose` | Enable debug logging |
 | `--help` | Show help |
@@ -304,11 +305,22 @@ The remote host must have `fastsync-server` available in `PATH`, or use
 working directory, so use a destination below that directory unless the
 remote server is otherwise configured with a matching authorized root.
 
+The remote `--stdio` server argv is composed by the client, so it must never
+be trusted to opt a root receiver into super-user activities: `--allow-super`
+is rejected with `--stdio` and super stays off on that path. Operators
+exposing `fastsync-server --stdio` over SSH must use a forced command (e.g. an
+`authorized_keys` `command=` entry) if the default must hold.
+
 ```bash
 ssh user@host 'mkdir -p destination'
 ./build/client /path/to/source user@host:destination
 ```
 
+FastSync is **push-only**: the source (first argument) is always a local
+directory and only the destination may be remote. A remote source such as
+`client user@host:src ./local` (a "pull") is intentionally not supported; see
+[RSYNC_COMPAT.md](RSYNC_COMPAT.md#direction).
+
 ### TCP transfer
 
 Start the FastSync server:
@@ -498,7 +510,8 @@ link-target transfer remains incomplete. |
 | `--ca <path>` | CA file for peer verification. |
 | `--destination-root <path>` | Confine received files to this server-side root;
 defaults to the current directory. |
-| `--allow-delete` | Permit client delete manifests. Deletion is refused by default. |
+| `--allow-delete` | Permit client delete manifests. Deletion is refused by default. This also gates `--force` (which can recursively replace/remove a destination directory tree). |
+| `--allow-super` | Standalone TCP listener only: keep super-user activities enabled for a **root** receiver. Without it a root standalone server forces `SUPER_MODE_OFF`, so client `--devices`/`--write-devices`/`--super` and client-chosen ownership requests are skipped/refused. Rejected with `--stdio` (the SSH remote argv is client-composed; use a forced command if the default must hold). No effect when not root. Daemon modules opt in per module with `client owner = yes`. |
 | `-v`, `--verbose` | Enable debug logging. |
 | `--help` | Print server usage. |
 

RSYNC_COMPAT.md

@@ -257,14 +257,14 @@ why plain `--append` works on the normal atomic path, not only with `--inplace`.
 | `-N`, `--crtimes` | Preserve create times | ⛔ Impossible/Divergence | Birth-times cannot be set by any portable filesystem call (`utimensat`/`futimens` only set atime/mtime), so this row is an explicit **Impossible/Divergence** (Phase 7 Wave B). Capture + transmit stays: `statx(STATX_BTIME)` on Linux records the source birth time as a wire field; the receiver logs a debug note that it cannot be applied and continues — never failing the transfer and never pretending it worked. On platforms without `statx` it parses as a documented no-op (flag accepted; nothing is captured). Implies metadata transmission. Wire: new `crtime` fields + a `preserve_crtimes` config boolean; `PROTOCOL_VERSION` bumped **2.11.0 → 2.12.0** (see the Phase-4 metadata-time notes) |
 | `-O`, `--omit-dir-times` | Omit dirs from --times | ✅ Implemented | Real modifier now that FastSync preserves directory times. With metadata on, the scanner captures every traversed source directory's mtime (and atime under `-U`) and the sender transmits them in trailing `STATUS_DIR_TIMES` frame(s) **after all file data and the optional delete manifest** (chunked at the receiver's `MAX_MANIFEST_ENTRIES` per-frame cap); a dir-time entry only RECORDS metadata and never creates the directory, so empty source directories stay untransferred. The receiver defers applying them until its delete / `--delay-updates` publication phases have committed, so writing or removing a child never clobbers a parent directory's mtime (rsync applies directory times at the end for exactly this reason). When `-O` is set (the boolean crosses the wire) the receiver does not apply any of them; without `-O` an `-a`/`--preserve` transfer now restores directory times (reversing the old "never preserves dir times" divergence). Wire change: the terminal `STATUS_DIR_TIMES` frame; `PROTOCOL_VERSION` bumped **2.16.0 → 2.17.0** |
 | `-J`, `--omit-link-times` | Omit symlinks from --times | ✅ Implemented | Real modifier now that FastSync preserves symlink times. Symlink entries already carried their metadata on `STATUS_SYMLINK`; the receiver now applies it with **no-follow primitives only** (`utimensat(..., AT_SYMLINK_NOFOLLOW)`, plus best-effort `fchmodat(..., AT_SYMLINK_NOFOLLOW)` and policy-gated `fchownat(..., AT_SYMLINK_NOFOLLOW)`), so the link itself is stamped without ever dereferencing it, confined fd-relative below the authorized receive root. A symlink has no children, so the times are applied immediately at creation. When `-J` is set (the boolean crosses the wire) the receiver skips the timestamps (mode/ownership are unaffected); without `-J` an `-a`/`-l` transfer restores symlink mtimes. Wire change alongside `-O`: the shared `STATUS_DIR_TIMES` frame; `PROTOCOL_VERSION` bumped **2.16.0 → 2.17.0** |
-| `--super` | Receiver attempts super-user activities | ✅ Implemented | Phase 7 Wave E: receiver-side **safe-subset + clear-refusal** privilege model, tri-state `super_mode` (auto/on/off). `--super` **permits** the receiver to attempt super-user activities — ownership application and char/block device-node creation — that are already confined fd-relative below the authorized receive root; `--no-super` **forbids** them even when the receiver is root; the default (`auto`) preserves the pre-existing **best-effort** behavior of *attempting* them (not only when already root: an unprivileged attempt is refused by the kernel and skipped per entry, matching FastSync's history). The server additionally accepts an operator-level `--no-super` veto that forces `OFF` for every connection it accepts (so it also refuses any client `--copy-as`/`--super`); the `--fake-super` owner replay and the `--write-devices` write path are gated by the same policy. **FastSync never elevates**: no `setuid`/`seteuid`/`setgid` is ever called, and `--super` never bypasses the confinement floor (`file_open_secure_parent`, `O_NOFOLLOW`, root checks) — it only permits an attempt that is already confined. `--super` does **not** imply `--numeric-ids` and never enables client-chosen ownership on its own: ownership is applied only when an explicit identity policy (`--usermap`/`--groupmap`/`--chown`/`--numeric-ids`/`--copy-as`) is also given. A non-root receiver given `--super` logs exactly one warning at activation and each confined attempt is then refused by the kernel and skipped per entry (never aborts); `--no-super` suppresses ownership, char/block `mknod`, `--write-devices` and the fake-super owner replay, while unprivileged FIFO creation is unaffected. Wire: one trailing `super_mode` int on the config frame (validated 0..2), sent **before** the `--copy-as` block (fixed order: super int, then copy-as presence int + ids); `PROTOCOL_VERSION` bumped **2.17.0 → 2.18.0**. **Documented divergence from rsync:** rsync's `--super` runs the receiver with elevated privilege; FastSync only permits a confined attempt and never elevates |
+| `--super` | Receiver attempts super-user activities | ✅ Implemented | Phase 7 Wave E: receiver-side **safe-subset + clear-refusal** privilege model, tri-state `super_mode` (auto/on/off). `--super` **permits** the receiver to attempt super-user activities — ownership application and char/block device-node creation — that are already confined fd-relative below the authorized receive root; `--no-super` **forbids** them even when the receiver is root; the default (`auto`) preserves the pre-existing **best-effort** behavior of *attempting* them (not only when already root: an unprivileged attempt is refused by the kernel and skipped per entry, matching FastSync's history). The server additionally accepts an operator-level `--no-super` veto that forces `OFF` for every connection it accepts (so it also refuses any client `--copy-as`/`--super`); a **privileged (root) standalone TCP listener now also defaults to `OFF`** unless the operator opts in with the new server-only `--allow-super` flag (the flag is **rejected with `--stdio`**, whose remote argv is composed by the client and must never defeat the secure default; operators exposing `fastsync-server --stdio` over SSH need a forced command if the default must hold.  An unprivileged receiver is unchanged, since the kernel refuses the confined attempts anyway; the `--daemon` path keeps its per-module `client owner = yes` opt-in); the `--fake-super` owner replay and the `--write-devices` write path are gated by the same policy. **FastSync never elevates**: no `setuid`/`seteuid`/`setgid` is ever called, and `--super` never bypasses the confinement floor (`file_open_secure_parent`, `O_NOFOLLOW`, root checks) — it only permits an attempt that is already confined. `--super` does **not** imply `--numeric-ids` and never enables client-chosen ownership on its own: ownership is applied only when an explicit identity policy (`--usermap`/`--groupmap`/`--chown`/`--numeric-ids`/`--copy-as`) is also given. A non-root receiver given `--super` logs exactly one warning at activation and each confined attempt is then refused by the kernel and skipped per entry (never aborts); `--no-super` suppresses ownership, char/block `mknod`, `--write-devices` and the fake-super owner replay, while unprivileged FIFO creation is unaffected. Wire: one trailing `super_mode` int on the config frame (validated 0..2), sent **before** the `--copy-as` block (fixed order: super int, then copy-as presence int + ids); `PROTOCOL_VERSION` bumped **2.17.0 → 2.18.0**. **Documented divergence from rsync:** rsync's `--super` runs the receiver with elevated privilege; FastSync only permits a confined attempt and never elevates |
 | `--fake-super` | Store/recover privileged attrs via xattrs | ✅ Implemented | Phase 7 Wave B: full record **and replay**. The receiver writes the source `uid:gid:mode:mtime_sec:mtime_nsec` into a reserved `user.fastsync.stat` xattr on each written file (best-effort, fd-relative, format unchanged), then immediately re-applies it via `fake_super_restore_fd`: `fchown` (only where privileged — a non-root EPERM/EACCES is skipped silently, matching FastSync's identity philosophy), `fchmod`, and `futimens`. The OWNER leg is additionally skipped unless an explicit ownership identity policy (`--numeric-ids`/`--usermap`/`--groupmap`/`--chown`/`--copy-as`) is active — `--fake-super` on its own only *records* the source owner and must not act as an un-gated chown primitive — when `--no-super` forbids super-user activities (even for root), or when an active `--copy-as` is authoritative, so the recorded source owner can never override a forced `--copy-as` owner; the xattr record is still stored/replayed for a later privileged restore and mode/mtime still apply, so unprivileged `--fake-super` keeps working. The restored mode goes through the same sanitization as the normal metadata path (group/other write bits are never granted, so a recorded 0666 restores as 0644), so fake-super replay can never grant group/other-write that plain `--preserve` would refuse. Absence or a malformed record is a silent no-op, never fatal. The recording format diverges from rsync's `user.rsync.%stat%`; no cross-tool conversion is attempted. Implies metadata transmission so the source uid/gid/mode/mtime are available. Both it and `-X`/`-A` are incompatible with `-s` (chunk serialization), rejected up front |
 | `--open-noatime` | Avoid changing access time when opening files | ✅ Implemented | Sender-side policy: the sender opens source files with `O_NOATIME` (Linux) when reading them for transfer, so the open/read does NOT bump the source's on-disk access time. Degrades safely when `O_NOATIME` is unavailable (not defined) or refused (`EPERM`, since it needs `CAP_FOWNER` or file ownership): the code falls back to a normal open, so the data always transfers — only the atime-bump is skipped. It does not itself capture/preserve atime; it only avoids modifying it. **Client-only, never crosses the wire.** Exposed as `file_open_for_read()` and applied to both the buffered data path and the sendfile path |
 | `--numeric-ids` | Do not map uid/gid by name | ✅ Implemented | Ownership is applied through FastSync's opt-in identity path (see the Phase-4 identity notes below). `--numeric-ids` is a mapping-policy modifier: when applying ownership it uses the transmitted numeric uid/gid directly, skipping the name lookup. Without an ownership-affecting option it is inert (FastSync only applies ownership when the user opts in). It does not need `-M` to be parsed, but ownership is only applied when metadata (hence the source uid/gid) is actually transmitted (see the notes) |
 | `--usermap=STRING` | Map usernames | ✅ Implemented | Opt-in ownership application. rsync subset implemented: comma-separated `FROM:TO` rules evaluated in order, first match wins; `FROM`/`TO` are group/user names (resolved on the SOURCE machine at parse time), `*` (FROM matches any id / TO = the receiving process's current euid), and an `@N` or bare `N` numeric id. Rules are carried over the wire as resolved numeric id pairs; the receiver applies a matching rule (else falls back to `--chown`, `--numeric-ids`, then a best-effort name lookup) via an fd-relative `fchown`. Malformed/unresolvable specs are rejected with a clear error, never a silent no-op. Implies metadata preservation so the source uid/gid travel. Only effective when the receiver can actually change ownership (root or membership); otherwise it warns and continues |
 | `--groupmap=STRING` | Map group names | ✅ Implemented | Same rsync subset and semantics as `--usermap` but for the group (gid) side and the group databases. See the Phase-4 identity notes |
 | `--chown=USER:GROUP` | Map owner and group | ✅ Implemented | Opt-in ownership override applied receiver-side. Forms: `USER:GROUP`, `USER` (owner only), `:GROUP` (group only); a `*` for USER/GROUP means the current/root user or group as appropriate; an `@N`/bare `N` numeric id is accepted. A `:` inside a name may be escaped as `\:`. Equivalent to a trailing `*:*` usermap+groupmap rule (so an explicit `--usermap`/`--groupmap` match wins). Malformed or unresolvable specs are clear parse errors. Implies metadata preservation. Only effective when the receiver has permission to chown; otherwise it warns and continues (rsync parity) |
-| `--copy-as=USER[:GROUP]` | Perform the copy as another user/group | ✅ Implemented | Safe-subset implementation, an explicit divergence from rsync's **real identity switching**. rsync makes the receiving process actually assume USER/GROUP (setuid/setgid); FastSync's receiver is multithreaded, so a real credential drop would be unsafe and is never attempted — FastSync never calls `setuid`/`seteuid`/`setgid`. Instead the receiver FORCES the ownership of every entry it writes to `copy_as_uid`/`copy_as_gid` through the existing confined, fd-relative identity path (the same `fchown`/`fchownat` mechanism as `--chown`/`--usermap`/`--groupmap`; symlinks use `fchownat(..., AT_SYMLINK_NOFOLLOW)`, and directories — including intermediate parents created implicitly while writing a nested file — and char/block/FIFO nodes are owned no-follow too, so a directory never keeps the receiver's owner while its children get the target owner), with `--copy-as` at the **highest priority** — it beats usermap/groupmap/`--chown`/`--numeric-ids` and the best-effort name lookup. This REQUIRES a privileged (root) receiver: an unprivileged receiver REFUSES the whole transfer up front at the config handshake (`server_module_gate`, running inside `config_receive_with_validate` before the `STATUS_OK` ack) with a clear error and no file data exchanged — never a silent wrong-ownership result. A server running with an operator `--no-super` veto also refuses it, and a **daemon** refuses `--copy-as`, like every other client-chosen-ownership request (`--numeric-ids`/`--chown`/`--usermap`/`--groupmap`/`--fake-super`/explicit `--super`), unless the selected module opts in with `client owner = yes`; without that per-module opt-in a daemon must not honor an arbitrary client-selected owner (the standalone listener and SSH `--stdio` server keep honoring these for their single operator-authorized root). `--fake-super` interaction: `--copy-as` is authoritative, so the recorded source owner is never replayed over the forced target owner. If the ownership apply still fails with EPERM/EACCES (capability-restricted root, root-squash, read-only mount) the failure is logged at ERROR and the **entry is reported as failed** rather than written with the wrong owner, which fails the transfer (fail-fast) so overall success is never reported with the wrong owner. USER is resolved on the client against the user database (a name, an `@N`/bare `N` numeric id, or `*` meaning the client's current euid); when `:GROUP` is present it is resolved against the group database (`*` meaning the client's egid). **Group-default rule:** when the group is omitted FastSync uses the user's primary gid (`getpwuid(uid)->pw_gid`); a numeric id with no local passwd entry has no primary gid to look up, so `gid` falls back to `uid` (documented divergence). Malformed/empty/unresolvable specs are clear parse errors, never a silent no-op. Never elevates privileges and never bypasses the confined receive root. Implies metadata preservation (the source uid/gid must be transmitted). Wire: a new trailing config-frame block **sent after** the `--super` int (presence int, then the two int32 ids, both validated `>= 0` on receive; the ids are also rejected if they do not fit int32 at CLI parse time); `PROTOCOL_VERSION` bumped **2.17.0 → 2.18.0** |
+| `--copy-as=USER[:GROUP]` | Perform the copy as another user/group | ✅ Implemented | Safe-subset implementation, an explicit divergence from rsync's **real identity switching**. rsync makes the receiving process actually assume USER/GROUP (setuid/setgid); FastSync's receiver is multithreaded, so a real credential drop would be unsafe and is never attempted — FastSync never calls `setuid`/`seteuid`/`setgid`. Instead the receiver FORCES the ownership of every entry it writes to `copy_as_uid`/`copy_as_gid` through the existing confined, fd-relative identity path (the same `fchown`/`fchownat` mechanism as `--chown`/`--usermap`/`--groupmap`; symlinks use `fchownat(..., AT_SYMLINK_NOFOLLOW)`, and directories — including intermediate parents created implicitly while writing a nested file — and char/block/FIFO nodes are owned no-follow too, so a directory never keeps the receiver's owner while its children get the target owner), with `--copy-as` at the **highest priority** — it beats usermap/groupmap/`--chown`/`--numeric-ids` and the best-effort name lookup. This REQUIRES a privileged (root) receiver: an unprivileged receiver REFUSES the whole transfer up front at the config handshake (`server_module_gate`, running inside `config_receive_with_validate` before the `STATUS_OK` ack) with a clear error and no file data exchanged — never a silent wrong-ownership result. A server running with an operator `--no-super` veto also refuses it; a privileged (root) standalone TCP listener refuses it by default too and only honors it after the operator passes `--allow-super` (the flag is rejected with `--stdio`, where the client-composed remote argv could otherwise defeat the default; a forced command is required if the default must hold), and a **daemon** refuses `--copy-as`, like every other client-chosen-ownership request (`--numeric-ids`/`--chown`/`--usermap`/`--groupmap`/`--fake-super`/explicit `--super`), unless the selected module opts in with `client owner = yes`; without that per-module opt-in a daemon must not honor an arbitrary client-selected owner (a root standalone listener honors these for its single operator-authorized root only when started with `--allow-super`). `--fake-super` interaction: `--copy-as` is authoritative, so the recorded source owner is never replayed over the forced target owner. If the ownership apply still fails with EPERM/EACCES (capability-restricted root, root-squash, read-only mount) the failure is logged at ERROR and the **entry is reported as failed** rather than written with the wrong owner, which fails the transfer (fail-fast) so overall success is never reported with the wrong owner. USER is resolved on the client against the user database (a name, an `@N`/bare `N` numeric id, or `*` meaning the client's current euid); when `:GROUP` is present it is resolved against the group database (`*` meaning the client's egid). **Group-default rule:** when the group is omitted FastSync uses the user's primary gid (`getpwuid(uid)->pw_gid`); a numeric id with no local passwd entry has no primary gid to look up, so `gid` falls back to `uid` (documented divergence). Malformed/empty/unresolvable specs are clear parse errors, never a silent no-op. Never elevates privileges and never bypasses the confined receive root. Implies metadata preservation (the source uid/gid must be transmitted). Wire: a new trailing config-frame block **sent after** the `--super` int (presence int, then the two int32 ids, both validated `>= 0` on receive; the ids are also rejected if they do not fit int32 at CLI parse time); `PROTOCOL_VERSION` bumped **2.17.0 → 2.18.0** |
 
 **Phase-4 metadata-time notes:** `-U/--atimes`, `-N/--crtimes`,
 `-O/--omit-dir-times`, `-J/--omit-link-times`, and `--open-noatime` are new.
@@ -639,8 +639,9 @@ now transmits targets (the prior behavior was broken/partial); its status moved
 - **Host access control (`hosts allow`/`hosts deny`):** both keys accept a comma- and/or whitespace-separated list of patterns and may appear globally and/or per module (multiple config-file lines append; a `--dparam` override replaces). Supported patterns are `*` (match all), an IPv4 or IPv6 literal (`10.0.0.1`, `2001:db8::1`), and an IPv4/IPv6 CIDR (`10.0.0.0/8`, `2001:db8::/32`). Hostname patterns are **not** supported: because the peer is always a numeric address and no reverse DNS is performed, a hostname/glob pattern would silently never match, so it is rejected at load time (fail-closed) instead of being accepted as a dead rule. An IPv4 peer on a dual-stack IPv6 listener is normalized from its `::ffff:a.b.c.d` form so IPv4 patterns match it. rsync-like semantics: a matching `hosts deny` rejects; if any `hosts allow` entries exist, a peer matching none of them is rejected; deny takes precedence over allow. The daemon enforces the global list first, then the selected module's list, **before authentication** in `server_module_gate`, with an audit log line naming the peer, the module and the outcome. The numeric peer address is obtained with `getpeername`+`inet_ntop` (`utils_fd_peer_ip`, handling both address families); when it cannot be obtained a module with any ACL fails closed (refused), while an ACL-free module continues and logs at debug. A malformed pattern (e.g. an out-of-range CIDR prefix) is a parse error at load time.
 - **Connection caps, shared registry and auth lockout:** the global `max connections` key (default 100) is plumbed into the listener (`transport_tcp.c`), which rejects a connection once the accept-loop parent's active-child count reaches it; the IPv4/IPv6 peer is logged for every accepted connection. Because the listener forks one child per connection, the per-module `max connections` cap, the global `max connections per host` cap, and the auth-failure counter live in a fixed-size registry carved from an anonymous shared mapping (`daemon_limits.c`, `mmap(MAP_SHARED|MAP_ANONYMOUS)`) created by the parent before the accept loop, so every forked child shares the same counters (C11 atomics only — never a pthread lock, which can deadlock in a forked child). The parent reserves a registry slot per accepted connection and the child records the selected module and source IP once known; the parent's `SIGCHLD` handler reclaims the slot when the child dies (including `SIGKILL`) and re-derives the per-module and per-source occupancy counts from the surviving REGISTERED slots, so a child killed mid-registration cannot leak a count. The per-source table has a bounded lifetime: an entry with no live connection is reclaimed after its lockout expires or it has been idle (300 s); if the table is genuinely full the per-source cap/lockout fails open for new sources (per-module cap and ACLs still apply) with a rate-limited warning. The per-module cap (0 = unlimited) is enforced after the module lookup and before auth; per-source identity reuses the normalized numeric peer address (`utils_fd_peer_ip`, IPv4-mapped IPv6 collapsed to IPv4), and a trusted loopback peer (127.0.0.0/8 / `::1`, `utils_fd_peer_is_local`) is exempt from the per-source cap and the auth lockout because all local clients share one address (the per-module/global caps still apply). Clients behind a shared NAT/proxy address likewise share one per-source budget and lockout counter. A failed authentication increments the shared per-source failure count and, once `auth lockout threshold` (default 10; 0 disables) is reached, the source is refused for `auth lockout duration` seconds (default 300) before any challenge is sent, even when the next attempt is handled by a different forked child; a successful authentication clears the counter. On a failed authentication the per-connection child still sleeps the global `auth failure delay` (default 500 ms, 0 disables, capped at 5000) via `nanosleep`, rate-limiting online guessing without delaying a success. A missing registry (allocation failure) degrades to the global cap and host ACLs rather than refusing to start.
 - **Module selection & confinement:** the client requests a module with an rsync-style `host::module[/path]` destination. The module name crosses the wire as a trailing string on the config frame (bumping `PROTOCOL_VERSION` 2.14.0 → 2.15.0; the bump is required because the config-frame layout changed and the strict same-version handshake is what prevents a peer from desynchronizing on the new trailing field). The daemon looks the module up in ITS OWN config and uses the module's `path` as the authorized root through the exact same `configure_authorization` confinement the standalone server applies to `--destination-root` (`file_open_secure_parent`, `has_path_traversal`, `path_is_within`); the client never supplies the root, every client-chosen-ownership/super-user request is refused unless the module declares `client owner = yes` (the daemon's per-module opt-in, see below), and the operator `--no-super` veto forces super-user activities off for every daemon connection. The client's `/path` part is relative inside the module and is rejected if absolute or if it contains `..`. Unknown modules are refused before any data moves (the run fails cleanly at the config handshake). An absolute destination and a module request against a non-daemon server are also refused.
-- **`client owner` (client-chosen-ownership opt-in):** by default a daemon module refuses every request that would let the client pick an owner or ask for super-user activities — `--numeric-ids`, `--chown`, `--usermap`/`--groupmap`, `--fake-super`, `--copy-as`, and an explicit `--super` — at the config handshake (before `STATUS_OK`), because a daemon has no per-module opt-in for client-chosen ownership and any anonymous client could otherwise force arbitrary owner ids inside the module root. `client owner = yes` opts a single module in, allowing those requests within that module's root (the standalone listener and the SSH `--stdio` server always honor them for their single operator-authorized root). Without the opt-in the daemon also forces super-user **device** activity off for that connection — char/block device-node creation (`--devices`) and `--write-devices` — even under the default `AUTO` mode, so a non-opted module can never be made to `mknod` or write a raw device; those entries are skipped (not refused) so an ordinary `-a` push still succeeds without device nodes. The opt-in does **not** lift the privilege requirement: `--copy-as` still needs a root receiver, and the operator `--no-super` veto still forces super-user activities off for every connection. The daemon logs a prominent startup warning for each `client owner = yes` module so the operator's deliberate choice is visible.
+- **`client owner` (client-chosen-ownership opt-in):** by default a daemon module refuses every request that would let the client pick an owner or ask for super-user activities — `--numeric-ids`, `--chown`, `--usermap`/`--groupmap`, `--fake-super`, `--copy-as`, and an explicit `--super` — at the config handshake (before `STATUS_OK`), because a daemon has no per-module opt-in for client-chosen ownership and any anonymous client could otherwise force arbitrary owner ids inside the module root. `client owner = yes` opts a single module in, allowing those requests within that module's root (a root standalone TCP listener honors them for its single operator-authorized root only when started with `--allow-super`; the flag is rejected with `--stdio`, whose client-composed remote argv must never opt back into super mode). Without the opt-in the daemon also forces super-user **device** activity off for that connection — char/block device-node creation (`--devices`) and `--write-devices` — even under the default `AUTO` mode, so a non-opted module can never be made to `mknod` or write a raw device; those entries are skipped (not refused) so an ordinary `-a` push still succeeds without device nodes. The opt-in does **not** lift the privilege requirement: `--copy-as` still needs a root receiver, and the operator `--no-super` veto still forces super-user activities off for every connection. The daemon logs a prominent startup warning for each `client owner = yes` module so the operator's deliberate choice is visible.
 - **`read only` safe default:** every network transfer FastSync currently supports is a push that writes under the module root, so a `read only` module refuses the connection (clear server log "module is read only"; the client exits non-zero, nothing is transferred). A future pull/list operation can be opened up when it exists; the knob is already stored.
+- **Direction — remote source / pull is intentionally unsupported:** FastSync is push-only. The first positional argument is always a **local** source directory and the second is the destination; only the destination is parsed for remote syntax (`user@host:path` SSH, `host::module[/path]` daemon). A remote source such as `fastsync user@host:src ./local` is deliberately **not** implemented: rsync has no pull flag (direction is positional), so supporting a remote source is an optional feature rather than a compatibility requirement, and it would require a protocol role reversal (server as sender, client as receiver) across both transports. FastSync documents this as an intentional limitation rather than a missing rsync option. <a id="direction"></a>
 - **`auth users` (A7 SCRAM-SHA-256 authentication):** a module that declares `auth users` requires the client to present credentials. The config frame carries ONLY the username; the daemon answers an auth-required module with `STATUS_AUTH_CHALLENGE` (PBKDF2 iteration count, 16-byte salt, 32-byte server nonce), the client answers with `STATUS_AUTH_RESPONSE` (fresh 32-byte client nonce + a 32-byte ClientProof), and the daemon accepts only when the proof verifies **and** the username is **on the module's `auth users` list** and has a store entry, replying `STATUS_AUTH_OK` with a 32-byte ServerSignature the client verifies before proceeding. Verification is constant-time over fixed 32-byte keys (the compare runs even for a miss), username membership uses a constant-time full-length scan, and an unknown/off-list user still receives a challenge and runs the same math against a dummy verifier: a deterministic per-username salt (`HMAC-SHA256(store dummy key, username)`), the store-wide uniform iteration count and dummy keys. Re-probing the same unknown username therefore yields an identical salt and iteration count while a different username yields a different salt, so there is no user-enumeration or timing oracle. The daemon logs the username but **never the password, proof or keys**. A module WITHOUT `auth users` stays open (legitimate rsync configuration); credentials sent to such a module are ignored. Read-only is orthogonal: even a correctly authenticated push to a `read only` module is still refused (all FastSync network transfers write). Fail-closed policy: a daemon whose config declares `auth users` on any module refuses to start unless a credential store was given (`--password-file` and/or `--early-input`); a missing or empty store is never silently treated as "open". A failed handshake (missing credentials, unknown/off-list user, wrong proof or malformed data) yields a single generic `STATUS_AUTH_FAILED` and the daemon closes before any data moves. The dummy key is persisted in an owner-only `<store_path>.dummykey` sidecar (auto-created on first load, mode 0600) so the dummy salt stays stable across daemon restarts, closing the restart-gated enumeration channel. The sidecar is secret material and must be protected like the credential store (owner-only 0600, included with the store in backups and rotation). It must be preserved across restarts for that guarantee; if it cannot be created (a process-substitution/FIFO store path such as `/dev/fd/N`, a read-only filesystem, a missing directory, or a create/write/fsync/link/fchmod failure), the daemon logs a warning and uses a transient per-run key, so unknown-user challenges change across restarts and the cross-restart guarantee does not hold for that deployment. One residual is accepted: the store iteration count is observable pre-auth by design, since the miss path must match a hit. **Transport policy (hardening A7-3/S1):** an auth-required module accepts credentials only when either (a) the connection is an encrypted, verified TLS connection whose client certificate matches `--client-cn`, or (b) the connection is plaintext from a loopback TCP peer **and** the operator explicitly passed `--allow-unauthenticated`. A remote plaintext peer, and a loopback plaintext peer without that flag, are refused at the config gate before any challenge is sent; `--allow-unauthenticated` never permits remote plaintext auth (remote peers still require verified TLS). Daemon modules are a `--daemon`-only feature — the SSH `--stdio` path never loads a daemon config and is not an auth transport for them. Because the loopback allowance trusts whichever peer the kernel reports as `127.0.0.1`, it assumes nothing relays remote connections to the daemon: a local TCP forwarder or TLS-terminating proxy in front of an auth-module listener makes remote clients appear as loopback and bypasses the mutual-TLS identity check, so do not front an auth-module listener with such a relay.
 - **Credential store format:** server `--password-file`/`--early-input` files are line-based `user:$fastsync$1$pbkdf2-sha256$<iters>$<salt_b64>$<stored_key_b64>$<server_key_b64>`, one per line (standard base64; 16-byte salt, 32-byte keys; `iters` in `[100000, 10000000]`, default 600000). Every entry in the resulting store must agree on `iters` (a store whose entries disagree, or where a layered `--early-input` disagrees with `--password-file`, is rejected). Generate lines with `fastsync-server --hash-credentials FILE [--iterations N]`; the emitted lines are secret material, so redirect them to an owner-only (mode 0600) file (the tool warns on stderr if stdout is a group/other-accessible regular file). Blank lines and lines starting with `#`/`;` are comments; the parser is strict (a malformed line fails the whole load, so a typo can never let a different set of users in). **The legacy `user:SHA256HEX` form is hard-rejected** with an actionable "legacy" error; there is no auto-upgrade, so a replayable bearer digest can never be loaded by a 2.19.0 daemon. The client `--password-file` holds `user:password` on its first meaningful line (the literal password, used only for the handshake then burned); keep both files readable only by their owner (mode 0600). Per-username wire length is bounded (256 chars) and every decoded salt/key length is validated. Loading the store also maintains an owner-only `<store_path>.dummykey` sidecar (auto-created, mode 0600, exactly 32 bytes) holding the store-wide dummy key that shapes unknown-user challenges; persist it across daemon restarts so those challenges stay stable, and treat a sidecar with the wrong owner, a mode other than exactly 0600, the wrong size or the wrong type as a fatal load error (fail closed). If the sidecar cannot be created (e.g. a process-substitution store path such as `/dev/fd/N`, a read-only filesystem, a missing directory, or a create/write/fsync/link/fchmod failure), the daemon logs a warning and uses a transient per-run key, so the cross-restart stability guarantee does not hold there.
 - **Plaintext caveat:** an auth-required module is refused, **before any challenge is sent**, unless the connection is encrypted and verified TLS whose client certificate matches the server's `--client-cn`, or it is plaintext from a loopback TCP peer **and** the operator passed `--allow-unauthenticated`. A remote plaintext peer, and a loopback plaintext peer without that flag, never receive a challenge, and `--allow-unauthenticated` never permits remote plaintext auth (remote peers still require verified TLS). On the loopback plaintext transport that remains permitted, a local sniffer could still read the challenge and response and mount an **offline dictionary attack** against a weak password, so use `--tls` for any real deployment. `--client-cn` matches the certificate CN only (not a subjectAltName), which is acceptable for a private CA. Clients sending daemon credentials with `--password-file` to a non-loopback daemon must use `--tls`; the client rejects such a destination before any network I/O. Unlike the old challenge-less exchange there is **no replay**: the proof is bound to the fresh per-connection server nonce, so a captured `STATUS_AUTH_RESPONSE` cannot be reused on another connection (an integration test proxies the daemon and proves this). TLS client-CN (`--client-cn`) is an independent transport identity check and composes with password auth; because `--tls` already mandates `--client-cn`, a TLS auth connection always verifies the client CN, so both checks necessarily apply together on such a connection.
@@ -662,7 +663,7 @@ now transmits targets (the prior behavior was broken/partial); its status moved
 | `--trust-sender` | Trust remote sender's file list | ✅ Implemented | Long-form-only, receiver-local policy that never crosses the wire. The receiver skips its redundant up-front re-validation of the incoming file list (empty/`..` path rejection and the escaping-symlink-target containment), trusting the sender instead of double-checking (fewer checks, faster, potentially unsafe, matching rsync). Off by default. The low-level fd-relative confinement primitives (`file_open_secure_parent`, the O_NOFOLLOW parent walk, leaf/destination confinement) are deliberately KEPT even under `--trust-sender`, so a hostile sender still cannot write or link outside the authorized root (see Phase-5 notes below) |
 | `--old-args` | Disable modern arg protection | ✅ Implemented | SSH-only; accepted for CLI compatibility but is now a **documented no-op**: FastSync always single-quote-escapes the remote server path and each `--remote-option` value (`ssh_build_remote_command`), so a metacharacter-bearing `--rsync-path` can never be interpreted by the remote shell. The flag no longer disables that quoting (the old raw-construction behavior was an injection foot-gun and is removed); the safety-relevant behavior is identical either way |
 | `--ignore-missing-args` | Ignore missing source args | ✅ Implemented | FastSync has a single source-root argument (which always exists), so the "explicitly requested source arguments" are the `--files-from` entries and the flags only ever apply there (inert without `--files-from`, like `-R`). Without the flag a listed-but-missing entry stays a hard pre-transfer error (nothing is transferred). With it each missing entry is skipped: nothing is sent for it, it never enters the keep-set, and the run succeeds for the rest — an all-missing non-empty list succeeds transferring nothing, matching rsync. `--dirs` + `--files-from` missing entries are skipped the same way. Every skipped entry is logged and a per-run warning names the count, so the handling is never a silent no-op. Divergences: an EMPTY `--files-from` file stays a hard error in every mode (no argument was requested at all; rsync likewise reports "no source files specified"); missing-arg skipping only applies to the pre-transfer list validation, so an entry that is present at preflight and vanishes mid-transfer still fails (matching rsync, whose flag "does not affect subsequent vanished-file errors"); `--no-ignore-missing-args` is not a supported negation |
-| `--delete-missing-args` | Delete missing source args | ✅ Implemented | Implies `--ignore-missing-args` (order-independent) and additionally removes each missing entry's destination mirror receiver-side. The mirror is computed exactly like a present sibling's wire path: the bare relative entry under `-R`, otherwise the full source-mirror path below the destination root. rsync parity, verified against the man page: it does **not** imply `--delete` generally and is "independent of any other type of delete processing" — unrelated destination extras are untouched unless `--delete` is also present. Composition with `--delete` + timing: the exact-path deletions commit with the manifest, early for `--delete-before`/`--delete-during`, else only after a fully-successful transfer (delete-after/commit). A non-empty directory mirror is removed only when `--force` or `--delete` is in effect (otherwise it is left with a warning and the run continues, like rsync); an absent mirror is a no-op. An explicitly listed missing arg is a user request, not an excluded file: its deletion is never blocked by the filter-exclusion protection of excluded destination mirrors (a mirror sitting inside a filter-excluded directory is still removed). Safety/policy: gated by the server `--allow-delete` policy like `--delete`; the request paths cross the wire only in the delete-manifest frame and are confined by the same receiver validation as the keep-set (non-empty, relative, traversal-free, bounded by the per-section/per-frame manifest caps); the `--delay-updates` staging directory and basis snapshots are protected exactly as in the extras walker. Divergence: the missing-args deletions are not counted toward `--max-delete` (they are explicit per-path requests, not discovered extras). See the Phase-3 wire note below for the `PROTOCOL_VERSION` bump |
+| `--delete-missing-args` | Delete missing source args | ✅ Implemented | Implies `--ignore-missing-args` (order-independent) and additionally removes each missing entry's destination mirror receiver-side. The mirror is computed exactly like a present sibling's wire path: the bare relative entry under `-R`, otherwise the full source-mirror path below the destination root. rsync parity, verified against the man page: it does **not** imply `--delete` generally and is "independent of any other type of delete processing" — unrelated destination extras are untouched unless `--delete` is also present. Composition with `--delete` + timing: the exact-path deletions commit with the manifest, early for `--delete-before`/`--delete-during`, else only after a fully-successful transfer (delete-after/commit). A non-empty directory mirror is removed only when `--force` or `--delete` is in effect (otherwise it is left with a warning and the run continues, like rsync); an absent mirror is a no-op. `--force` is deletion authority and is therefore gated by the server `--allow-delete` policy exactly like `--delete`/`--delete-missing-args`: without it the receiver clears the flag, so a client cannot use `--force` to recursively replace or remove a destination directory tree. An explicitly listed missing arg is a user request, not an excluded file: its deletion is never blocked by the filter-exclusion protection of excluded destination mirrors (a mirror sitting inside a filter-excluded directory is still removed). Safety/policy: gated by the server `--allow-delete` policy like `--delete`; the request paths cross the wire only in the delete-manifest frame and are confined by the same receiver validation as the keep-set (non-empty, relative, traversal-free, bounded by the per-section/per-frame manifest caps); the `--delay-updates` staging directory and basis snapshots are protected exactly as in the extras walker. Divergence: the missing-args deletions are not counted toward `--max-delete` (they are explicit per-path requests, not discovered extras). See the Phase-3 wire note below for the `PROTOCOL_VERSION` bump |
 
 ## 16. Batch Operations
 
@@ -832,9 +833,9 @@ These are the last compatibility items and the closing phase toward rsync flag p
 
 **Wave E (LAST) — Privilege: `--super`/`--no-super` and `--copy-as=USER[:GROUP]` (✅ implemented).** FastSync adopts a **safe-subset + clear-refusal** privilege model: it never blind-elevates and never calls `setuid`/`seteuid`/`setgid`. All privileged operations remain fd-relative and confined below the authorized receive root.
 
-`--super`/`--no-super` set a receiver-side tri-state `Config->super_mode` (`SUPER_MODE_AUTO`/`ON`/`OFF`). `privilege_super_permitted()` / `privilege_super_mode_permitted()` (src/shared/identity.c) return true for `ON` and `AUTO` (AUTO preserves FastSync's historical best-effort attempt, where the kernel refuses an unprivileged call and the caller skips it) and false only for `OFF`. The gate covers every super-user activity FastSync performs: ownership application (`identity_apply_ownership`/`_link`), char/block device-node creation (`file_save_special_to_disk`), writes into an existing device (`--write-devices`), and the `--fake-super` owner replay. Unprivileged FIFO creation is deliberately unaffected. `--super` does **not** imply `--numeric-ids`: ownership is applied only when an explicit identity policy (`--usermap`/`--groupmap`/`--chown`/`--numeric-ids`/`--copy-as`) is also given. `--no-super` suppresses those activities even for a root receiver. A non-root receiver given `--super` logs one warning at activation (`identity_set_active`); each confined attempt is then refused by the kernel and skipped, never aborting. The confinement floor is unchanged (`file_open_secure_parent`, `O_NOFOLLOW`, root/path checks). Operator control: the server CLI accepts `--no-super`, a veto that forces `OFF` for every connection, refuses any client `--copy-as`, and neutralizes an explicit `--super` (the connection is accepted but no super-user activity is attempted). On a daemon, a module that has not opted in with `client owner = yes` additionally has super-user device activity forced off (see the Daemon Mode notes).
+`--super`/`--no-super` set a receiver-side tri-state `Config->super_mode` (`SUPER_MODE_AUTO`/`ON`/`OFF`). `privilege_super_permitted()` / `privilege_super_mode_permitted()` (src/shared/identity.c) return true for `ON` and `AUTO` (AUTO preserves FastSync's historical best-effort attempt, where the kernel refuses an unprivileged call and the caller skips it) and false only for `OFF`. The gate covers every super-user activity FastSync performs: ownership application (`identity_apply_ownership`/`_link`), char/block device-node creation (`file_save_special_to_disk`), writes into an existing device (`--write-devices`), and the `--fake-super` owner replay. Unprivileged FIFO creation is deliberately unaffected. `--super` does **not** imply `--numeric-ids`: ownership is applied only when an explicit identity policy (`--usermap`/`--groupmap`/`--chown`/`--numeric-ids`/`--copy-as`) is also given. `--no-super` suppresses those activities even for a root receiver. A non-root receiver given `--super` logs one warning at activation (`identity_set_active`); each confined attempt is then refused by the kernel and skipped, never aborting. The confinement floor is unchanged (`file_open_secure_parent`, `O_NOFOLLOW`, root/path checks). Operator control: the server CLI accepts `--no-super`, a veto that forces `OFF` for every connection, refuses any client `--copy-as`, and neutralizes an explicit `--super` (the connection is accepted but no super-user activity is attempted). A privileged (root) standalone TCP listener instead defaults to `OFF` and requires the server-only `--allow-super` opt-in to attempt any super-user activity (the flag is rejected with `--stdio`, whose client-composed remote argv must never defeat the default; use a forced command if the default must hold); a non-root server is unchanged. On a daemon, a module that has not opted in with `client owner = yes` additionally has super-user device activity forced off (see the Daemon Mode notes).
 
-`--copy-as=USER[:GROUP]` is the safe subset. FastSync's receiver is multithreaded, so a real credential switch is unsafe; instead the receiver forces the ownership of **every entry it writes** — regular files, symlinks, directories (including implicitly-created parents), and special nodes — to the resolved target ids through the confined fd-relative identity path. USER is resolved on the client (name, `@N`/bare N, or `*` = client euid); when `:GROUP` is omitted the user's primary gid is used (falling back to `gid == uid` for a numeric id with no local passwd entry). It requires a privileged (root) receiver: an unprivileged receiver refuses the whole transfer at the config handshake, before `STATUS_OK`, so no data is ever written with the wrong ownership. A `--copy-as` chown failure on a capability-restricted root is logged at ERROR (never silently downgraded). `--copy-as` implies metadata (`--no-preserve` is rejected) and `--fake-super` cannot override it. Daemon policy: a `--daemon` receiver refuses **every** client-chosen-ownership / super-user request — `--numeric-ids`, `--chown`, `--usermap`/`--groupmap`, `--fake-super`, `--copy-as`, and explicit `--super` — unless the selected module opts in with `client owner = yes`; without that per-module opt-in any client could force arbitrary ownership inside the module root (the standalone listener and the SSH-launched `--stdio` server, which each serve one operator-authorized root, honor these requests). A `--copy-as` chown failure on a capability-restricted root marks the entry as failed rather than reporting success with the wrong owner.
+`--copy-as=USER[:GROUP]` is the safe subset. FastSync's receiver is multithreaded, so a real credential switch is unsafe; instead the receiver forces the ownership of **every entry it writes** — regular files, symlinks, directories (including implicitly-created parents), and special nodes — to the resolved target ids through the confined fd-relative identity path. USER is resolved on the client (name, `@N`/bare N, or `*` = client euid); when `:GROUP` is omitted the user's primary gid is used (falling back to `gid == uid` for a numeric id with no local passwd entry). It requires a privileged (root) receiver: an unprivileged receiver refuses the whole transfer at the config handshake, before `STATUS_OK`, so no data is ever written with the wrong ownership. A `--copy-as` chown failure on a capability-restricted root is logged at ERROR (never silently downgraded). `--copy-as` implies metadata (`--no-preserve` is rejected) and `--fake-super` cannot override it. Daemon policy: a `--daemon` receiver refuses **every** client-chosen-ownership / super-user request — `--numeric-ids`, `--chown`, `--usermap`/`--groupmap`, `--fake-super`, `--copy-as`, and explicit `--super` — unless the selected module opts in with `client owner = yes`; without that per-module opt-in any client could force arbitrary ownership inside the module root (a root standalone TCP listener, which serves one operator-authorized root, honors these requests only when started with `--allow-super`; the flag is rejected with `--stdio`). A `--copy-as` chown failure on a capability-restricted root marks the entry as failed rather than reporting success with the wrong owner.
 
 **Wire:** two trailing config-frame blocks after the `--iconv` spec, in fixed order — `send_privilege_options`/`receive_privilege_options` (one `super_mode` int, validated `0..2`), then `send_copy_as_options`/`receive_copy_as_options` (presence int + two int32 ids, validated `>= 0`, with `copy_as_set ⇒ use_metadata`). `PROTOCOL_VERSION` bumped **2.17.0 → 2.18.0**. **Divergences from rsync:** rsync's `--super` elevates the receiver and `--copy-as` actually switches its credentials; FastSync never elevates and only permits/forwards confined attempts, and `--copy-as` forces ownership rather than switching identity.
 

benchmark/bench.py

@@ -3,19 +3,24 @@
 
 Compares FastSync configs against rsync (no compression) and rsync+zstd.
 Data is ~75% random/incompressible and ~25% structured/compressible by default,
-controllable via --random-ratio.
+controllable via --random-ratio. Transfers are verified by default (source and
+destination must match) so a fast-but-broken copy is never counted.
 
 Usage:
     python3 benchmark/bench.py
     python3 benchmark/bench.py --runs 5 --profiles lan wan
     python3 benchmark/bench.py --random-ratio 0.5 --size-mb 50
     python3 benchmark/bench.py --delay 50ms --jitter 10ms --throughput 100mbit
+    python3 benchmark/bench.py --warm --runs 3
     python3 benchmark/bench.py --output json
 """
 import argparse
+import filecmp
 import json
+import math
 import os
 import random
+import shlex
 import shutil
 import socket
 import statistics
@@ -25,7 +30,10 @@ import tempfile
 import time
 
 PROJECT_ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
-BUILD_DIR = os.path.join(PROJECT_ROOT, "build")
+DEFAULT_BUILD_DIR = "build-bench"
+# Populated by configure_build_dirs(); default to the dedicated bench dir so
+# importing this module never depends on the user's existing build/ tree.
+BUILD_DIR = os.path.join(PROJECT_ROOT, DEFAULT_BUILD_DIR)
 SERVER_CMD = [os.path.join(BUILD_DIR, "server"), "--allow-unauthenticated"]
 CLIENT_CMD = [os.path.join(BUILD_DIR, "client")]
 BENCH_DIR = os.path.join(PROJECT_ROOT, "bench_data")
@@ -57,6 +65,7 @@ RSYNC_CONFIGS = [
     {"name": "rsync -z --zstd",    "flags": ["-z", "--zc", "zstd"],"tool": "rsync"},
 ]
 
+
 class RsyncDaemon:
     """Manages an rsync daemon for network-fair benchmarking."""
 
@@ -118,6 +127,12 @@ STRUCTURED_FILES = {
     "nested/another.txt": b"another nested file\n" * 50,
 }
 
+# Repeated text used to synthesize genuinely compressible filler of any size.
+COMPRESSIBLE_TEXT = (
+    b"FastSync benchmark payload: the quick brown fox jumps over the lazy dog. "
+    b"0123456789 ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz\n"
+)
+
 
 class Progress:
     """Simple progress bar with ETA."""
@@ -152,35 +167,127 @@ class Progress:
             sys.stderr.flush()
 
 
+def write_compressible(path, nbytes):
+    """Write exactly nbytes of highly compressible, repeated text content."""
+    if nbytes <= 0:
+        return
+    block = COMPRESSIBLE_TEXT * (max(1, 8192 // len(COMPRESSIBLE_TEXT)) + 1)
+    remaining = nbytes
+    with open(path, "wb") as f:
+        while remaining > 0:
+            piece = block if remaining >= len(block) else block[:remaining]
+            f.write(piece)
+            remaining -= len(piece)
+
+
 def generate_bench_data(source_dir, size_mb=25, random_ratio=0.75):
-    """Generate test data. ~random_ratio is incompressible, rest is structured."""
+    """Generate test data honouring the requested random/compressible split.
+
+    Exactly ``random_ratio * target`` bytes are incompressible random data and
+    the remainder is genuinely compressible structured/repeated content. The
+    measured byte counts are returned so callers can report the real mix.
+    """
     if os.path.exists(source_dir):
         shutil.rmtree(source_dir)
     os.makedirs(source_dir)
 
     target = size_mb * 1024 * 1024
-    structured_budget = int(target * (1 - random_ratio))
-    written = 0
+    random_budget = int(target * random_ratio)
+    compressible_budget = target - random_budget
+    compressible_written = 0
+    random_written = 0
+    files = 0
 
+    # A handful of fixed, human-meaningful files (directories, small files, a
+    # binary blob) as long as they fit inside the compressible budget.
     for rel_path, content in STRUCTURED_FILES.items():
-        if written >= structured_budget:
+        if compressible_written + len(content) > compressible_budget:
             break
         full_path = os.path.join(source_dir, rel_path)
         os.makedirs(os.path.dirname(full_path), exist_ok=True)
         with open(full_path, "wb") as f:
             f.write(content)
-        written += len(content)
+        compressible_written += len(content)
+        files += 1
 
-    os.makedirs(os.path.join(source_dir, "bulk"), exist_ok=True)
-    i = 0
-    while written < target:
-        chunk_size = min(5 * 1024 * 1024, target - written)
-        with open(os.path.join(source_dir, f"bulk/file_{i}.dat"), "wb") as f:
-            f.write(random.randbytes(chunk_size))
-        written += chunk_size
-        i += 1
+    # Fill the rest of the compressible share with generated repeated content.
+    if compressible_written < compressible_budget:
+        os.makedirs(os.path.join(source_dir, "compressible"), exist_ok=True)
+        i = 0
+        while compressible_written < compressible_budget:
+            chunk = min(1024 * 1024, compressible_budget - compressible_written)
+            write_compressible(os.path.join(source_dir, "compressible", f"text_{i}.dat"), chunk)
+            compressible_written += chunk
+            files += 1
+            i += 1
 
-    return written
+    # Incompressible share.
+    if random_written < random_budget:
+        os.makedirs(os.path.join(source_dir, "bulk"), exist_ok=True)
+        i = 0
+        while random_written < random_budget:
+            chunk = min(5 * 1024 * 1024, random_budget - random_written)
+            with open(os.path.join(source_dir, "bulk", f"file_{i}.dat"), "wb") as f:
+                f.write(random.randbytes(chunk))
+            random_written += chunk
+            files += 1
+            i += 1
+
+    return {
+        "total_bytes": compressible_written + random_written,
+        "compressible_bytes": compressible_written,
+        "random_bytes": random_written,
+        "files": files,
+    }
+
+
+def list_relative_files(root):
+    """Return the set of file paths (relative to root) under a directory."""
+    found = set()
+    for dirpath, _dirnames, filenames in os.walk(root):
+        for name in filenames:
+            full = os.path.join(dirpath, name)
+            found.add(os.path.relpath(full, root))
+    return found
+
+
+def verify_transfer(source_dir, dest_dir):
+    """Recursively check dest matches source (paths, sizes, content).
+
+    Returns (ok, detail). Content is compared byte-for-byte, never hashed, so
+    collisions are impossible. This is intentionally not part of the timing.
+    """
+    if not os.path.isdir(dest_dir):
+        return False, "destination directory missing"
+    src_files = list_relative_files(source_dir)
+    dst_files = list_relative_files(dest_dir)
+    if src_files != dst_files:
+        missing = src_files - dst_files
+        extra = dst_files - src_files
+        return False, f"path set mismatch (missing {len(missing)}, extra {len(extra)})"
+    for rel in sorted(src_files):
+        src = os.path.join(source_dir, rel)
+        dst = os.path.join(dest_dir, rel)
+        if os.path.getsize(src) != os.path.getsize(dst):
+            return False, f"size mismatch: {rel}"
+        if not filecmp.cmp(src, dst, shallow=False):
+            return False, f"content mismatch: {rel}"
+    return True, ""
+
+
+def percentile(values, pct):
+    """Linear-interpolation percentile (matches numpy's default method)."""
+    if not values:
+        return None
+    ordered = sorted(values)
+    if len(ordered) == 1:
+        return ordered[0]
+    rank = (len(ordered) - 1) * (pct / 100.0)
+    low = math.floor(rank)
+    high = math.ceil(rank)
+    if low == high:
+        return ordered[int(rank)]
+    return ordered[low] + (ordered[high] - ordered[low]) * (rank - low)
 
 
 def find_free_port():
@@ -208,18 +315,45 @@ def wait_proc(proc, timeout=5):
         proc.wait()
 
 
+def _tc_base_cmd():
+    """Return the command prefix for tc, honouring root vs sudo."""
+    tc = shutil.which("tc")
+    if not tc:
+        raise RuntimeError(
+            "tc (iproute2) not found in PATH; install iproute2 to use network profiles")
+    if os.geteuid() == 0:
+        return [tc]
+    sudo = shutil.which("sudo")
+    if sudo:
+        return [sudo, tc]
+    raise RuntimeError(
+        "applying network limits requires root or sudo; "
+        "re-run as root or install sudo")
+
+
+def _run_tc(args, check=True):
+    return subprocess.run(_tc_base_cmd() + args, check=check, capture_output=True)
+
+
 def netem_apply(delay=None, jitter=None, throughput=None, loss=None):
     """Apply tc/netem rules to loopback. Pass None to skip a parameter."""
     netem_reset()
-    cmd = ["sudo", "tc", "qdisc", "add", "dev", "lo", "root", "netem"]
+    params = []
     if throughput:
-        cmd += ["rate", throughput]
+        params += ["rate", throughput]
     if delay:
-        cmd += ["delay", delay, jitter or "0ms"]
+        params += ["delay", delay, jitter or "0ms"]
     if loss:
-        cmd += ["loss", loss]
-    if len(cmd) > 6:
-        subprocess.run(cmd, check=True, capture_output=True)
+        params += ["loss", loss]
+    if not params:
+        return
+    try:
+        _run_tc(["qdisc", "add", "dev", "lo", "root", "netem"] + params)
+    except subprocess.CalledProcessError as exc:
+        detail = exc.stderr.decode(errors="replace").strip() if exc.stderr else str(exc)
+        raise RuntimeError(f"failed to apply network profile via tc/netem: {detail}") from exc
+    except RuntimeError:
+        raise
 
 
 def netem_apply_profile(profile_name):
@@ -236,7 +370,11 @@ def netem_apply_profile(profile_name):
 
 
 def netem_reset():
-    subprocess.run("sudo tc qdisc del dev lo root".split(), capture_output=True)
+    """Best-effort removal of any loopback qdisc. Always safe to call."""
+    try:
+        _run_tc(["qdisc", "del", "dev", "lo", "root"], check=False)
+    except Exception:
+        pass
 
 
 def run_fastsync(source_dir, dest_dir, flags, port):
@@ -287,7 +425,81 @@ def run_transfer(config, source_dir, dest_dir, port=None, rsync_daemon=None):
         return run_fastsync(source_dir, dest_dir, config["flags"], port)
 
 
-def run_benchmark(source_dir, dest_dir, configs, runs, profile_name, progress=None):
+def apply_incremental_changes(source_dir, target_bytes):
+    """Add and modify a few files so a warm transfer has real work to do.
+
+    Returns a mutation record (changed byte count plus enough data to revert
+    and re-apply it) so every warm run can start from a pristine source.
+    """
+    modified_n = 3
+    added_n = 2
+    per_file = max(4096, target_bytes // (modified_n + added_n))
+    modified = {}
+    added = {}
+    changed = 0
+
+    existing = sorted(list_relative_files(source_dir))
+    if existing:
+        step = max(1, len(existing) // modified_n)
+        for rel in existing[::step][:modified_n]:
+            path = os.path.join(source_dir, rel)
+            original_size = os.path.getsize(path)
+            with open(path, "ab") as f:
+                f.write(random.randbytes(per_file))
+            modified[rel] = (original_size, per_file)
+            changed += per_file
+
+    for i in range(added_n):
+        os.makedirs(os.path.join(source_dir, "incremental"), exist_ok=True)
+        rel = os.path.join("incremental", f"new_{i}.dat")
+        write_compressible(os.path.join(source_dir, rel), per_file)
+        added[rel] = per_file
+        changed += per_file
+
+    return {"changed": changed, "modified": modified, "added": added}
+
+
+def revert_incremental_changes(source_dir, mutation):
+    """Undo apply_incremental_changes so the source is pristine again."""
+    if not mutation:
+        return
+    for rel, (original_size, _appended) in mutation["modified"].items():
+        path = os.path.join(source_dir, rel)
+        if os.path.exists(path):
+            with open(path, "r+b") as f:
+                f.truncate(original_size)
+    for rel in mutation["added"]:
+        path = os.path.join(source_dir, rel)
+        if os.path.exists(path):
+            os.remove(path)
+
+
+def reapply_incremental_changes(source_dir, mutation):
+    """Re-apply a mutation after an untimed pristine seed transfer."""
+    if not mutation:
+        return
+    for rel, (_original_size, appended) in mutation["modified"].items():
+        with open(os.path.join(source_dir, rel), "ab") as f:
+            f.write(random.randbytes(appended))
+    for rel, size in mutation["added"].items():
+        write_compressible(os.path.join(source_dir, rel), size)
+
+
+def expected_received_root(dest_dir, source_dir, tool):
+    """Where a tool places transferred files inside dest_dir.
+
+    FastSync mirrors the absolute source path under dest_dir (see the
+    integration suite's get_dest_received_dir); rsync copies the source tree
+    contents directly into dest_dir.
+    """
+    if tool == "rsync":
+        return dest_dir
+    return os.path.join(dest_dir, os.path.abspath(source_dir).lstrip(os.sep))
+
+
+def run_benchmark(source_dir, dest_dir, configs, runs, profile_name,
+                  measure_bytes, verify=True, warm=False, mutation=None,
+                  progress=None):
     """Run benchmark for all configs, returns list of results."""
     is_limited = profile_name != "unlimited"
     has_rsync = any(c["tool"] == "rsync" for c in configs)
@@ -303,7 +515,10 @@ def run_benchmark(source_dir, dest_dir, configs, runs, profile_name, progress=No
         results = []
         for config in configs:
             times = []
+            invalid = 0
             for run_idx in range(runs):
+                if warm:
+                    revert_incremental_changes(source_dir, mutation)
                 if os.path.exists(dest_dir):
                     shutil.rmtree(dest_dir)
                 os.makedirs(dest_dir, exist_ok=True)
@@ -311,15 +526,33 @@ def run_benchmark(source_dir, dest_dir, configs, runs, profile_name, progress=No
                 port = find_free_port()
                 server = None
                 try:
-                    if config["tool"] == "fastsync":
+                    if config["tool"] == "fastsync" or warm:
                         server = subprocess.Popen(
                             SERVER_CMD + ["-p", str(port)],
                             stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
                         )
                         wait_for_port(port)
 
+                    if warm:
+                        seed = run_transfer(config, source_dir, dest_dir, port, rsync_daemon)
+                        if seed is None:
+                            invalid += 1
+                            sys.stderr.write("    warm-mode seeding failed; run not counted\n")
+                            continue
+                        reapply_incremental_changes(source_dir, mutation)
+
                     t = run_transfer(config, source_dir, dest_dir, port, rsync_daemon)
-                    if t is not None:
+                    if t is None:
+                        invalid += 1
+                    elif verify:
+                        root = expected_received_root(dest_dir, source_dir, config["tool"])
+                        ok, detail = verify_transfer(source_dir, root)
+                        if ok:
+                            times.append(t)
+                        else:
+                            invalid += 1
+                            sys.stderr.write(f"    verification FAILED ({detail}); run not counted\n")
+                    else:
                         times.append(t)
                 finally:
                     if server:
@@ -332,15 +565,22 @@ def run_benchmark(source_dir, dest_dir, configs, runs, profile_name, progress=No
                 "config": config["name"],
                 "tool": config["tool"],
                 "profile": profile_name,
+                "warm": warm,
                 "runs": len(times),
+                "invalid": invalid,
                 "times": [round(t, 4) for t in times],
             }
             if times:
-                entry["p50"] = round(statistics.median(times), 4)
-                entry["p95"] = round(sorted(times)[int(len(times) * 0.95)], 4) if len(times) > 1 else entry["p50"]
+                p50 = percentile(times, 50)
+                p95 = percentile(times, 95)
+                entry["p50"] = round(p50, 4)
+                entry["p95"] = round(p95, 4)
                 entry["min"] = round(min(times), 4)
                 entry["max"] = round(max(times), 4)
                 entry["stdev"] = round(statistics.stdev(times), 4) if len(times) > 1 else 0.0
+                if measure_bytes:
+                    entry["throughput_mbps"] = round(
+                        (measure_bytes / (1024 * 1024)) / p50, 3)
             results.append(entry)
         return results
     finally:
@@ -350,44 +590,59 @@ def run_benchmark(source_dir, dest_dir, configs, runs, profile_name, progress=No
             netem_reset()
 
 
-def print_table(results, total_bytes, random_ratio):
+def print_table(results, measure_bytes, stats, warm):
     """Print results as a human-readable table grouped by profile."""
     profiles = {}
     for r in results:
         profiles.setdefault(r["profile"], []).append(r)
 
+    total = stats["total_bytes"]
+    comp_pct = stats["compressible_bytes"] / total * 100 if total else 0
+    rand_pct = stats["random_bytes"] / total * 100 if total else 0
+
     for profile, entries in profiles.items():
         params = NETWORK_PROFILES.get(profile, {})
-        print(f"\n{'=' * 85}")
+        print(f"\n{'=' * 95}")
         print(f" Profile: {profile.upper()}")
         if params.get("rate"):
             print(f" Network: {params['rate']}, {params['delay']} +/- {params['jitter']}, loss {params['loss']}")
         else:
             print(f" Network: unlimited")
-        print(f" Data: {total_bytes / (1024*1024):.1f} MB  ({random_ratio*100:.0f}% random, {(1-random_ratio)*100:.0f}% compressible)")
-        print(f"{'=' * 85}")
+        print(f" Data: {total / (1024*1024):.1f} MB  "
+              f"({rand_pct:.0f}% random, {comp_pct:.0f}% compressible actual)")
+        if warm:
+            print(f" Mode: warm (incremental) — measured {measure_bytes / (1024*1024):.2f} MB "
+                  f"changed after an untimed full seed")
+        else:
+            print(" Mode: cold (full copy)")
+        print(f"{'=' * 95}")
 
         fs_entries = [e for e in entries if e.get("tool") == "fastsync"]
         rsync_entries = [e for e in entries if e.get("tool") == "rsync"]
 
+        header = (f" {'Config':<38} {'p50':>8} {'p95':>8} {'min':>8} {'max':>8} "
+                  f"{'stdev':>8} {'MB/s':>9} {'runs':>5} {'bad':>4}")
+        rule = (f" {'-' * 38} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 8} "
+                f"{'-' * 8} {'-' * 9} {'-' * 5} {'-' * 4}")
+
         if fs_entries:
             print(f"\n FastSync:")
-            print(f" {'Config':<38} {'p50':>8} {'p95':>8} {'min':>8} {'max':>8} {'stdev':>8} {'runs':>5}")
-            print(f" {'-' * 38} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 5}")
+            print(header)
+            print(rule)
             for e in sorted(fs_entries, key=lambda x: x.get("p50", 999)):
                 _print_entry(e)
 
         if rsync_entries:
             print(f"\n rsync:")
-            print(f" {'Config':<38} {'p50':>8} {'p95':>8} {'min':>8} {'max':>8} {'stdev':>8} {'runs':>5}")
-            print(f" {'-' * 38} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 8} {'-' * 5}")
+            print(header)
+            print(rule)
             for e in sorted(rsync_entries, key=lambda x: x.get("p50", 999)):
                 _print_entry(e)
 
         if params.get("rate_bps") and fs_entries and rsync_entries:
             fs_best = min((e["p50"] for e in fs_entries if "p50" in e), default=None)
             rsync_best = min((e["p50"] for e in rsync_entries if "p50" in e), default=None)
-            theoretical = total_bytes / params["rate_bps"]
+            theoretical = measure_bytes / params["rate_bps"]
             if fs_best and rsync_best:
                 print(f"\n Theoretical max (line rate): {theoretical:.4f}s")
                 print(f" FastSync best:   {fs_best:.4f}s  ({theoretical/fs_best:.2f}x vs line rate)")
@@ -397,10 +652,43 @@ def print_table(results, total_bytes, random_ratio):
 
 def _print_entry(e):
     if "p50" in e:
+        tp = f"{e['throughput_mbps']:.2f}" if "throughput_mbps" in e else "N/A"
         print(f" {e['config']:<38} {e['p50']:>7.4f}s {e['p95']:>7.4f}s "
-              f"{e['min']:>7.4f}s {e['max']:>7.4f}s {e['stdev']:>7.4f} {e['runs']:>5}")
+              f"{e['min']:>7.4f}s {e['max']:>7.4f}s {e['stdev']:>7.4f} "
+              f"{tp:>9} {e['runs']:>5} {e.get('invalid', 0):>4}")
     else:
-        print(f" {e['config']:<38} {'N/A':>8} {'N/A':>8} {'N/A':>8} {'N/A':>8} {'N/A':>8} {e['runs']:>5}")
+        print(f" {e['config']:<38} {'N/A':>8} {'N/A':>8} {'N/A':>8} {'N/A':>8} "
+              f"{'N/A':>8} {'N/A':>9} {e['runs']:>5} {e.get('invalid', 0):>4}")
+
+
+def configure_build_dirs(build_dir):
+    """Install the selected build directory and derived binary paths."""
+    global BUILD_DIR, SERVER_CMD, CLIENT_CMD
+    if not os.path.isabs(build_dir):
+        build_dir = os.path.join(PROJECT_ROOT, build_dir)
+    BUILD_DIR = os.path.abspath(build_dir)
+    SERVER_CMD = [os.path.join(BUILD_DIR, "server"), "--allow-unauthenticated"]
+    CLIENT_CMD = [os.path.join(BUILD_DIR, "client")]
+
+
+def build_project():
+    """Configure (Release) and build into the dedicated bench build dir."""
+    if shutil.which("cmake") is None:
+        sys.stderr.write("cmake not found in PATH; cannot build\n")
+        sys.exit(1)
+    os.makedirs(BUILD_DIR, exist_ok=True)
+    configure = ["cmake", "-B", BUILD_DIR, "-S", PROJECT_ROOT,
+                 "-DCMAKE_BUILD_TYPE=Release"]
+    result = subprocess.run(configure, capture_output=True, text=True)
+    if result.returncode != 0:
+        sys.stderr.write("CMake configure failed:\n" + result.stdout + result.stderr + "\n")
+        sys.exit(1)
+    jobs = str(os.cpu_count() or 1)
+    result = subprocess.run(["cmake", "--build", BUILD_DIR, "-j", jobs],
+                            capture_output=True, text=True)
+    if result.returncode != 0:
+        sys.stderr.write("Build failed:\n" + result.stdout + result.stderr + "\n")
+        sys.exit(1)
 
 
 def main():
@@ -417,12 +705,20 @@ Custom network limits (--delay/--jitter/--throughput) override profiles.
 
 Data mix:
   Default is ~75%% random/incompressible + ~25%% structured/compressible,
-  reflecting typical real-world file sets.
+  reflecting typical real-world file sets. The actual mix is measured and
+  reported. Transfers are verified (destination must match source) unless
+  --no-verify is given.
+
+Warm mode:
+  --warm seeds the destination with an untimed full copy of a pristine base,
+  then measures only the incremental transfer after modifying a few files.
 
 Examples:
   %(prog)s --profiles wan --runs 5
   %(prog)s --throughput 50mbit --delay 30ms --jitter 5ms
   %(prog)s --random-ratio 0.5 --size-mb 100
+  %(prog)s --warm --runs 3 --no-rsync
+  %(prog)s --dry-run --size-mb 4 --random-ratio 0.25
 """)
     parser.add_argument("--runs", type=int, default=3,
                         help="Number of runs per config (default: 3)")
@@ -430,7 +726,8 @@ Examples:
                         choices=list(NETWORK_PROFILES.keys()),
                         help="Predefined network profiles (default: unlimited)")
     parser.add_argument("--configs", nargs="+", default=None,
-                        help="Custom FastSync config flags")
+                        help="Custom FastSync config flags (shell-quoted, e.g. "
+                             "\"-j -z --chunk-serialization\")")
     parser.add_argument("--size-mb", type=int, default=25,
                         help="Test data size in MB (default: 25)")
     parser.add_argument("--random-ratio", type=float, default=0.75,
@@ -445,6 +742,14 @@ Examples:
                         help="Custom packet loss (e.g. 1%%)")
     parser.add_argument("--no-rsync", action="store_true",
                         help="Skip rsync comparison")
+    parser.add_argument("--no-verify", action="store_true",
+                        help="Skip source/destination verification after each run")
+    parser.add_argument("--warm", action="store_true",
+                        help="Incremental mode: seed dest first, measure only changes")
+    parser.add_argument("--build-dir", default=DEFAULT_BUILD_DIR,
+                        help=f"Build directory (default: {DEFAULT_BUILD_DIR})")
+    parser.add_argument("--dry-run", action="store_true",
+                        help="Only generate data and report its composition, then exit")
     parser.add_argument("--progress", action="store_true",
                         help="Show progress bar with ETA")
     parser.add_argument("--output", choices=["table", "json"], default="table",
@@ -453,14 +758,48 @@ Examples:
                         help="Don't clean up test data")
     args = parser.parse_args()
 
+    if not 0.0 <= args.random_ratio <= 1.0:
+        parser.error("--random-ratio must be between 0.0 and 1.0")
+    if args.size_mb <= 0:
+        parser.error("--size-mb must be positive")
+
+    configure_build_dirs(args.build_dir)
+
+    # Generate data
+    source_dir = os.path.join(BENCH_DIR, "source")
+    dest_dir = os.path.join(BENCH_DIR, "dest")
+    stats = generate_bench_data(source_dir, args.size_mb, args.random_ratio)
+    total_bytes = stats["total_bytes"]
+    comp_pct = stats["compressible_bytes"] / total_bytes * 100 if total_bytes else 0
+    rand_pct = stats["random_bytes"] / total_bytes * 100 if total_bytes else 0
+    print(f"Generated {total_bytes / (1024*1024):.1f} MB in {stats['files']} files  "
+          f"({rand_pct:.0f}% random, {comp_pct:.0f}% compressible actual)",
+          file=sys.stderr)
+
+    if args.dry_run:
+        print(f"size_mb={args.size_mb} random_ratio={args.random_ratio:.4f} "
+              f"total_bytes={stats['total_bytes']} "
+              f"compressible_bytes={stats['compressible_bytes']} "
+              f"random_bytes={stats['random_bytes']} files={stats['files']}")
+        if not args.keep_data:
+            shutil.rmtree(BENCH_DIR, ignore_errors=True)
+        return
+
+    # Warm mode: keep a pristine base copy, then mutate the live source.
+    base_dir = None
+    measure_bytes = total_bytes
+    mutation = None
+    if args.warm:
+        change_target = max(64 * 1024, min(int(total_bytes * 0.01), 4 * 1024 * 1024))
+        mutation = apply_incremental_changes(source_dir, change_target)
+        measure_bytes = mutation["changed"]
+        revert_incremental_changes(source_dir, mutation)
+        print(f"Warm mode: each run seeds a full copy, then measures "
+              f"{measure_bytes / (1024*1024):.3f} MB of add/change deltas", file=sys.stderr)
+
     # Build (Release: benchmarking a debug build is meaningless)
-    print("Building (Release)...", file=sys.stderr)
-    configure = (f"cmake -B {BUILD_DIR} -S {PROJECT_ROOT} "
-                 f"-DCMAKE_BUILD_TYPE=Release > /dev/null 2>&1")
-    if os.system(configure) != 0:
-        print("CMake configure failed", file=sys.stderr); sys.exit(1)
-    if os.system(f"cmake --build {BUILD_DIR} -j$(nproc) > /dev/null 2>&1") != 0:
-        print("Build failed", file=sys.stderr); sys.exit(1)
+    print(f"Building (Release) into {BUILD_DIR}...", file=sys.stderr)
+    build_project()
 
     # Determine active profile for display
     has_custom_net = args.delay or args.jitter or args.throughput or args.loss
@@ -480,19 +819,10 @@ Examples:
     else:
         profiles_to_run = args.profiles or ["unlimited"]
 
-    # Generate data
-    source_dir = os.path.join(BENCH_DIR, "source")
-    dest_dir = os.path.join(BENCH_DIR, "dest")
-    total_bytes = generate_bench_data(source_dir, args.size_mb, args.random_ratio)
-    compressible_pct = (1 - args.random_ratio) * 100
-    random_pct = args.random_ratio * 100
-    print(f"Generated {total_bytes / (1024*1024):.1f} MB  "
-          f"({random_pct:.0f}% random, {compressible_pct:.0f}% compressible)",
-          file=sys.stderr)
-
-    # Build config list
+    # Build config list (shlex so quoted/space-separated flags survive)
     if args.configs:
-        fastsync_configs = [{"name": c, "flags": c.split(), "tool": "fastsync"} for c in args.configs]
+        fastsync_configs = [{"name": c, "flags": shlex.split(c), "tool": "fastsync"}
+                            for c in args.configs]
     else:
         fastsync_configs = list(FASTSYNC_CONFIGS)
 
@@ -509,9 +839,16 @@ Examples:
     all_results = []
     try:
         for profile in profiles_to_run:
-            results = run_benchmark(source_dir, dest_dir, configs, args.runs, profile, progress)
+            results = run_benchmark(source_dir, dest_dir, configs, args.runs, profile,
+                                    measure_bytes, verify=not args.no_verify,
+                                    warm=args.warm, mutation=mutation,
+                                    progress=progress)
             all_results.extend(results)
+    except RuntimeError as exc:
+        sys.stderr.write(f"error: {exc}\n")
+        sys.exit(1)
     finally:
+        netem_reset()
         if not args.keep_data:
             shutil.rmtree(BENCH_DIR, ignore_errors=True)
 
@@ -519,7 +856,7 @@ Examples:
     if args.output == "json":
         print(json.dumps(all_results, indent=2))
     else:
-        print_table(all_results, total_bytes, args.random_ratio)
+        print_table(all_results, measure_bytes, stats, args.warm)
         print()
 
 

shell.nix

@@ -3,11 +3,35 @@
 }:
 
 pkgs.mkShell {
+  # Development shell for FastSync.  Provides the host-side toolchain needed to
+  # build, lint, unit-test, integration-test and benchmark the project.
+  # It deliberately does NOT build on entry: run the CMake commands in README.md
+  # (or use the CI Docker image for exact CI parity).
   nativeBuildInputs = with pkgs; [
+    # build
     gcc
     cmake
     gnumake
     pkg-config
+    # lint / static analysis (matches CI)
+    clang-tools # clang-format
+    cppcheck
+    # tests
+    (python3.withPackages (ps: with ps; [ pytest pytest-xdist psutil ]))
+    openssh # SSH transport integration tests
+    # debugging
+    gdb
+    valgrind
+    # coverage
+    lcov
+    # benchmark tooling
+    rsync
+    iproute2 # tc/netem for network shaping
+    # misc
+    git
+    curl
+    nodejs
+    nixpkgs-fmt
     docker
     tea
   ];
@@ -15,14 +39,21 @@ pkgs.mkShell {
   buildInputs = with pkgs; [
     zstd
     openssl
-    (python3.withPackages (ps: with ps; [ pytest ]))
   ];
 
+  # The CMake configure step fetches xxHash via FetchContent, which needs
+  # network access; NIX_ENFORCE_PURITY must be off so the sandbox does not block.
   NIX_ENFORCE_PURITY = 0;
 
   shellHook = ''
     export NIX_ENFORCE_PURITY=0
-    cmake -B build
-    export PATH="$PWD/build:$PATH"
+    # Make an existing build tree available on PATH, but never build here.
+    if [ -d "$PWD/build" ]; then
+      export PATH="$PWD/build:$PATH"
+    fi
+    echo "FastSync dev shell ready."
+    echo "  Build:      cmake -B build -S . && cmake --build build -j\$(nproc)"
+    echo "  Unit:       ./build/tests"
+    echo "  CI parity:  docker run --rm --user \"\$(id -u):\$(id -g)\" -v \"\$PWD:/workspace\" -w /workspace gitea.tap-tap.win/taptap/fastsync-ci:v10 ..."
   '';
 }

src/client/client_cli.c

@@ -19,6 +19,7 @@
 #include "usage.h"
 #include "utils.h"
 #include <errno.h>
+#include <fcntl.h>
 #include <limits.h>
 #include <time.h>
 #include <signal.h>
@@ -27,6 +28,7 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <unistd.h>
 
 /* Async-signal-safe abort flag set by the SIGINT/SIGTERM handler.  Exposed via
  * client_send.h so the send loops can poll it.  Defined here (not in
@@ -429,8 +431,14 @@ static int parse_info_flags(const char* value, Config* config) {
   return 0;
 }
 
-/* Parse a string as an unsigned long long. Returns 0 on success, -1 on error. */
+/* Parse a string as an unsigned long long. Returns 0 on success, -1 on error.
+ * A leading '-'/'+' (or whitespace) is rejected outright: strtoull would
+ * otherwise silently wrap a negative value to a huge unsigned one. */
 static int parse_ull_arg(const char* val, unsigned long long* out, const char* optname) {
+  if (!val || val[0] < '0' || val[0] > '9') {
+    log_message(LOG_LEVEL_ERROR, "%s must be a non-negative integer", optname);
+    return -1;
+  }
   char* end;
   errno = 0;
   unsigned long long v = strtoull(val, &end, 10);
@@ -537,7 +545,10 @@ static int config_add_filter(Config* config, const char* rule) {
   char err[160];
   FilterRule* parsed = filter_rule_parse(rule, err, sizeof(err));
   if (!parsed) {
-    log_message(LOG_LEVEL_ERROR, "invalid --filter rule '%s': %s", rule, err);
+    char* escaped = output_escape(rule, log_get_8_bit_output());
+    log_message(LOG_LEVEL_ERROR, "invalid --filter rule '%s': %s",
+                escaped ? escaped : "<allocation failed>", err);
+    free(escaped);
     return -1;
   }
   filter_rule_free(parsed);
@@ -1303,10 +1314,15 @@ static bool cli_handle_ssh_and_pattern_options(CliParseCtx* ctx) {
       ctx->exit_code = -1;
       return true;
     }
-    if (val >= DELTA_MIN_FILE_SIZE)
+    if (val >= DELTA_MIN_FILE_SIZE && val <= DELTA_MAX_FILE_SIZE) {
       config->delta_max_file_size = val;
-    else
+    } else if (val < DELTA_MIN_FILE_SIZE) {
       log_message(LOG_LEVEL_WARNING, "--delta-max value %llu too small, using default", val);
+    } else {
+      log_message(LOG_LEVEL_ERROR, "--delta-max must not exceed %llu bytes",
+                  (unsigned long long)DELTA_MAX_FILE_SIZE);
+      ctx->exit_code = -1;
+    }
     return true;
   }
   return false;
@@ -1463,6 +1479,12 @@ static bool cli_handle_io_options(CliParseCtx* ctx) {
       ctx->exit_code = -1;
       return true;
     }
+    if (val > MAX_CHUNK_SIZE) {
+      log_message(LOG_LEVEL_ERROR, "--chunk-size must be between 1 and %llu",
+                  (unsigned long long)MAX_CHUNK_SIZE);
+      ctx->exit_code = -1;
+      return true;
+    }
     config->chunk_size = val;
     return true;
   }
@@ -1479,11 +1501,20 @@ static bool cli_handle_io_options(CliParseCtx* ctx) {
       fclose(config->log_file);
       config->log_file = NULL;
     }
-    FILE* lf = fopen(ctx->argv[++ctx->i], "a");
+    const char* log_path = ctx->argv[++ctx->i];
+    /* Refuse a symlinked target and never leak the descriptor across exec: an
+     * attacker who can plant a symlink in the working directory must not be
+     * able to redirect (or truncate) an arbitrary file via --log-file.  The log
+     * is created with owner-only permissions. */
+    int log_fd = open(log_path, O_WRONLY | O_CREAT | O_APPEND | O_NOFOLLOW | O_CLOEXEC, 0600);
+    FILE* lf = log_fd >= 0 ? fdopen(log_fd, "a") : NULL;
     if (!lf) {
-      char* escaped = output_escape(ctx->argv[ctx->i], false);
+      int open_errno = errno;
+      if (log_fd >= 0)
+        close(log_fd);
+      char* escaped = output_escape(log_path, false);
       log_message(LOG_LEVEL_ERROR, "could not open log file '%s': %s",
-                  escaped ? escaped : "<allocation failed>", strerror(errno));
+                  escaped ? escaped : "<allocation failed>", strerror(open_errno));
       free(escaped);
       ctx->exit_code = -1;
       return true;
@@ -2011,8 +2042,27 @@ static int read_patterns_from_file(const char* filepath, char*** patterns, int*
   }
   char* line = NULL;
   size_t line_size = 0;
-  ssize_t n;
-  while ((n = getline(&line, &line_size, fp)) != -1) {
+  while (true) {
+    ssize_t n = utils_getdelim_bounded(fp, &line, &line_size, '\n', UTILS_MAX_LINE_LEN);
+    if (n < 0) {
+      /* output_escape() may allocate (and clobber errno): capture the reader's
+       * errno first so an over-long line is still reported as EFBIG. */
+      int saved_errno = errno;
+      char* escaped = output_escape(filepath, false);
+      if (saved_errno == EFBIG) {
+        log_message(LOG_LEVEL_ERROR, "pattern file '%s' has a line exceeding %d bytes",
+                    escaped ? escaped : "<allocation failed>", (int)UTILS_MAX_LINE_LEN);
+      } else {
+        log_message(LOG_LEVEL_ERROR, "could not read pattern file '%s': %s",
+                    escaped ? escaped : "<allocation failed>", strerror(saved_errno));
+      }
+      free(escaped);
+      free(line);
+      fclose(fp);
+      return -1;
+    }
+    if (n == 0)
+      break;
     char* p = line;
     while (*p == ' ' || *p == '\t')
       p++;

src/client/client_send.c

@@ -310,8 +310,11 @@ static bool files_from_list_check(const Config* config, ArrayList* missing_dest,
     return false;
   }
   if (set->count == 0) {
+    char* escaped_list =
+        output_escape(config->files_from ? config->files_from : "", log_get_8_bit_output());
     log_message(LOG_LEVEL_ERROR, "--files-from file '%s' contains no entries; nothing to transfer",
-                config->files_from ? config->files_from : "");
+                escaped_list ? escaped_list : "<allocation failed>");
+    free(escaped_list);
     return false;
   }
   bool ignore = config->ignore_missing_args || config->delete_missing_args;
@@ -329,7 +332,10 @@ static bool files_from_list_check(const Config* config, ArrayList* missing_dest,
       free(full);
       if (ignore) {
         (*skipped_out)++;
-        log_info_message(LOG_INFO_MISC, "skipping missing --files-from entry '%s'", entry);
+        char* escaped_entry = output_escape(entry, log_get_8_bit_output());
+        log_info_message(LOG_INFO_MISC, "skipping missing --files-from entry '%s'",
+                         escaped_entry ? escaped_entry : "<allocation failed>");
+        free(escaped_entry);
         if (config->delete_missing_args && missing_dest) {
           char* mirror = files_from_missing_dest_path(config, entry);
           if (!mirror || !array_list_add(missing_dest, mirror)) {
@@ -340,8 +346,13 @@ static bool files_from_list_check(const Config* config, ArrayList* missing_dest,
         }
         continue;
       }
-      log_message(LOG_LEVEL_ERROR, "--files-from entry '%s' not found in source '%s'", entry,
-                  config->send_directory);
+      char* escaped_entry = output_escape(entry, log_get_8_bit_output());
+      char* escaped_src = output_escape(config->send_directory, log_get_8_bit_output());
+      log_message(LOG_LEVEL_ERROR, "--files-from entry '%s' not found in source '%s'",
+                  escaped_entry ? escaped_entry : "<allocation failed>",
+                  escaped_src ? escaped_src : "<allocation failed>");
+      free(escaped_entry);
+      free(escaped_src);
       return false;
     }
     free(full);
@@ -588,8 +599,12 @@ static void remove_transferred_sources(const Config* config, ArrayList* paths) {
       close(dirfd);
       continue;
     }
-    if (unlinkat(dirfd, leaf, 0) != 0)
-      log_message(LOG_LEVEL_WARNING, "Could not remove source file %s", source->path);
+    if (unlinkat(dirfd, leaf, 0) != 0) {
+      char* escaped_path = output_escape(source->path, log_get_8_bit_output());
+      log_message(LOG_LEVEL_WARNING, "Could not remove source file %s",
+                  escaped_path ? escaped_path : "<allocation failed>");
+      free(escaped_path);
+    }
     close(dirfd);
   }
 }
@@ -2120,7 +2135,7 @@ int write_batch_from_source(const Config* config, const char* batch_path) {
     prepared_scanner_destroy(&prepared);
     return 1;
   }
-  int fd = open(batch_path, O_WRONLY | O_CREAT | O_TRUNC, 0644);
+  int fd = open(batch_path, O_WRONLY | O_CREAT | O_TRUNC | O_NOFOLLOW | O_CLOEXEC, 0600);
   if (fd < 0) {
     log_perror("could not create batch file");
     directory_scanner_destroy(scanner);
@@ -2135,8 +2150,10 @@ int write_batch_from_source(const Config* config, const char* batch_path) {
       if (f == NULL || f->data == NULL)
         continue;
       if (f->data->size > 0 && f->data->data == NULL && !file_load_data(f)) {
+        char* escaped_path = output_escape(f->path ? f->path : "", log_get_8_bit_output());
         log_message(LOG_LEVEL_ERROR, "batch: failed to load data for %s",
-                    f->path ? f->path : "<no path>");
+                    escaped_path ? escaped_path : "<allocation failed>");
+        free(escaped_path);
         ok = false;
         break;
       }

src/client/client_validation.c

@@ -25,12 +25,14 @@ bool validate_config(const Config* config) {
   /* A dry-run of a local batch apply is not meaningful: --read-batch bypasses
      the client-side scan/server decision entirely, so dry-run would have no
      wire state to report (and must not be used as a mutation escape hatch).
-     --only-write-batch likewise never contacts a receiver.  Reject both up
+     --only-write-batch likewise never contacts a receiver.  --write-batch DOES
+     run a live transfer but additionally mutates the filesystem by emitting the
+     batch file, so a dry-run must not write it either.  Reject all three up
      front instead of silently ignoring --dry-run. */
-  if (config->dry_run && (read_batch || only_write_batch)) {
+  if (config->dry_run && (read_batch || only_write_batch || write_batch)) {
     log_message(LOG_LEVEL_ERROR,
-                "--dry-run cannot be combined with --read-batch or --only-write-batch; "
-                "a dry-run of a local batch apply is not meaningful");
+                "--dry-run cannot be combined with --read-batch, --only-write-batch, or "
+                "--write-batch; a dry-run must not mutate anything, including batch files");
     return false;
   }
   if (read_batch) {

src/client/scanner.c

@@ -179,7 +179,7 @@ static bool entry_passes_selection(const FileListSet* file_list, const FilterRul
 static void scanner_capture_xattrs(const DirectoryScanner* scanner, File* file) {
   if (!scanner || !file || !(scanner->options.preserve_xattrs || scanner->options.preserve_acls))
     return;
-  file->xattrs = xattr_capture_path(file->path);
+  file->xattrs = xattr_capture_path(file->path, scanner->options.preserve_acls);
 }
 
 /* Apply --hard-links (-H) detection to one regular File.  On a sibling (a
@@ -284,7 +284,10 @@ static int open_directory_filter_context(DirectoryScanner* scanner, const Filter
       filter_file_read(scanner->current_path, scanner->current_rel ? scanner->current_rel : "",
                        &exists, err, sizeof(err));
   if (!own) {
-    log_message(LOG_LEVEL_ERROR, "invalid .rsync-filter in %s: %s", scanner->current_path, err);
+    char* escaped_path = output_escape(scanner->current_path, log_get_8_bit_output());
+    log_message(LOG_LEVEL_ERROR, "invalid .rsync-filter in %s: %s",
+                escaped_path ? escaped_path : "<allocation failed>", err);
+    free(escaped_path);
     scanner->failed = true;
     return -1;
   }
@@ -777,11 +780,19 @@ static File* dirs_file_for_entry(DirectoryScanner* scanner, const char* entry) {
        nothing (missing entries never appear there).  Without the flags it stays
        a hard pre-transfer error. */
     if (scanner->options.ignore_missing_args) {
-      log_info_message(LOG_INFO_MISC, "skipping missing --files-from entry '%s'", entry);
+      char* escaped_entry = output_escape(entry, log_get_8_bit_output());
+      log_info_message(LOG_INFO_MISC, "skipping missing --files-from entry '%s'",
+                       escaped_entry ? escaped_entry : "<allocation failed>");
+      free(escaped_entry);
       free(abs_path);
       return NULL;
     }
-    log_message(LOG_LEVEL_ERROR, "--dirs listed entry is not present under the source: %s", entry);
+    {
+      char* escaped_entry = output_escape(entry, log_get_8_bit_output());
+      log_message(LOG_LEVEL_ERROR, "--dirs listed entry is not present under the source: %s",
+                  escaped_entry ? escaped_entry : "<allocation failed>");
+      free(escaped_entry);
+    }
     free(abs_path);
     scanner->failed = true;
     return NULL;
@@ -1434,7 +1445,7 @@ static void scan_root_entry(const ScannerOptions* options, const FilterNode* roo
   }
   if ((options->preserve_xattrs || options->preserve_acls) &&
       !(file->link_group != 0 && !file->link_first))
-    file->xattrs = xattr_capture_path(file->path);
+    file->xattrs = xattr_capture_path(file->path, options->preserve_acls);
   if (!array_list_add(root_files, file)) {
     free(rel);
     file_destroy(file);

src/server/server.c

@@ -37,6 +37,15 @@ static bool allow_unauthenticated;
  * root), so no super-user activity is attempted and any client --copy-as is
  * refused.  Set once in main before the accept loop / stdio handler. */
 static bool server_no_super;
+/* --allow-super: locally-launched standalone TCP opt-in that preserves the
+ * historical permissive super mode for a root receiver.  When false, a
+ * privileged standalone receiver forces SUPER_MODE_OFF for every connection
+ * (C3), so a client cannot make it create device nodes / write raw devices /
+ * apply client-chosen ownership.  It is REJECTED for --stdio (the SSH remote
+ * argv is composed by the client, so it must never be able to opt a root
+ * receiver back into super mode); the --stdio path always keeps the secure
+ * default. */
+static bool server_allow_super;
 static const char* required_client_cn;
 /* --iconv CONVERT_SPEC the server was itself started with (borrowed argv
  * pointer).  Its LOCAL half may override the local charset the client assumed;
@@ -187,13 +196,25 @@ static bool tls_client_identity_allowed(SSL* ssl) {
   X509* certificate = SSL_get1_peer_certificate(ssl);
   if (!certificate)
     return false;
-  char common_name[256];
-  int length = X509_NAME_get_text_by_NID(X509_get_subject_name(certificate), NID_commonName,
-                                         common_name, sizeof(common_name));
   size_t required_length = strlen(required_client_cn);
-  bool allowed = length >= 0 && (size_t)length == required_length &&
-                 required_length < sizeof(common_name) &&
-                 credentials_secure_equal(common_name, required_client_cn, required_length);
+  bool allowed = false;
+  X509_NAME* subject = X509_get_subject_name(certificate);
+  int index = subject ? X509_NAME_get_index_by_NID(subject, NID_commonName, -1) : -1;
+  if (index >= 0) {
+    X509_NAME_ENTRY* entry = X509_NAME_get_entry(subject, index);
+    ASN1_STRING* data = entry ? X509_NAME_ENTRY_get_data(entry) : NULL;
+    /* Convert the CN to UTF-8 to get its FULL byte length: unlike
+     * X509_NAME_get_text_by_NID (which truncates an over-long CN to the buffer
+     * and reports the truncated length), ASN1_STRING_to_UTF8 never truncates, so
+     * an exactly-required-length CN is accepted while an over-long one cannot be
+     * prefix-matched by a shorter required name. */
+    unsigned char* utf8 = NULL;
+    int cn_length = data ? ASN1_STRING_to_UTF8(&utf8, data) : -1;
+    if (cn_length >= 0 && (size_t)cn_length == required_length)
+      allowed = credentials_secure_equal((const char*)utf8, required_client_cn, required_length);
+    if (utf8)
+      OPENSSL_free(utf8);
+  }
   X509_free(certificate);
   return allowed;
 }
@@ -592,6 +613,22 @@ static const char* server_module_gate(const Config* config, void* context) {
     if (gate_ctx)
       gate_ctx->super_mode_override = SUPER_MODE_OFF;
   }
+  /* C3: a privileged (root) STANDALONE receiver defaults to SUPER_MODE_OFF.
+   * Without this a client --devices/--write-devices/--super would let a root
+   * server create arbitrary device nodes and write raw devices, and
+   * client-chosen ownership (--numeric-ids/--chown/--usermap/--groupmap) would
+   * be applied, with no operator opt-in.  The operator must pass --allow-super
+   * to restore the historical permissive behavior; the flag is rejected for
+   * --stdio, whose client-composed argv must never defeat this default (an
+   * operator exposing `fastsync-server --stdio` over SSH needs a forced command
+   * to keep the permissive behavior).  An unprivileged receiver is unaffected
+   * (the kernel refuses the confined attempts) and the daemon path keeps its
+   * per-module `client owner = yes` gate. */
+  if (g_daemon_conf == NULL && geteuid() == 0 && !server_allow_super) {
+    effective.super_mode = SUPER_MODE_OFF;
+    if (gate_ctx)
+      gate_ctx->super_mode_override = SUPER_MODE_OFF;
+  }
   /* --copy-as (P7 Wave E, protocol 2.18.0): FastSync's safe subset forces the
      ownership of every written entry to the requested ids, which needs a
      privileged (root) receiver.  An unprivileged receiver REFUSES the whole
@@ -751,6 +788,13 @@ void handler(int file_descriptor) {
     goto done;
   }
   config->use_delete = config->use_delete && allow_delete;
+  /* --force (receiver-side) is deletion authority too: it lets an incoming
+   * regular file recursively remove a non-empty destination directory tree, and
+   * lets --delete-missing-args remove a non-empty directory mirror.  Without
+   * the operator's --allow-delete it must be inert, exactly like --delete and
+   * --delete-missing-args, so a client cannot use --force to bypass the delete
+   * policy. */
+  config->force_delete = config->force_delete && allow_delete;
   /* --iconv (protocol 2.16.0): install the receiver-side wire->local conversion
      now that the client's full CONVERT_SPEC has been received and validated,
      before any received file name is decoded.  The server's own --iconv (if
@@ -1010,6 +1054,13 @@ static void print_server_usage(void) {
   printf("  --no-super          Operator veto: never attempt super-user activities\n");
   printf("                      (ownership, device nodes) even as root, and refuse\n");
   printf("                      any client --copy-as/--super request\n");
+  printf("  --allow-super       Standalone TCP listener only: keep super-user\n");
+  printf("                      activities enabled for a root receiver.  Without it a\n");
+  printf("                      root standalone server forces SUPER_MODE_OFF, so client\n");
+  printf("                      --devices/--write-devices/--super and ownership\n");
+  printf("                      requests are refused/skipped.  Never honored with\n");
+  printf("                      --stdio (the SSH remote argv is client-composed, so\n");
+  printf("                      super stays off there); no effect when not root\n");
   printf("  --iconv=LOCAL[,REMOTE]  Declare this server's LOCAL charset for file-name\n");
   printf("                      conversion: received names are translated to this\n");
   printf("                      charset (the wire charset still comes from the\n");
@@ -1134,6 +1185,9 @@ int main(int argc, char* argv[]) {
   trust_sender = opts.trust_sender;
   allow_unauthenticated = opts.allow_unauthenticated;
   server_no_super = opts.no_super;
+  /* --stdio rejects --allow-super at parse time; force it off here as well so
+   * this process-global policy cannot be re-enabled by a future caller. */
+  server_allow_super = opts.allow_super && !opts.stdio_mode;
   server_iconv_spec = opts.iconv_spec;
   signal(SIGINT, cleanup);
   signal(SIGTERM, cleanup);

src/server/server_cli.c

@@ -179,6 +179,8 @@ int server_cli_parse(int argc, char* argv[], ServerCliOptions* opts, char* err,
       opts->trust_sender = true;
     } else if (arg_is(argv[i], "--no-super")) {
       opts->no_super = true;
+    } else if (arg_is(argv[i], "--allow-super")) {
+      opts->allow_super = true;
     } else if (arg_is(argv[i], "--allow-unauthenticated")) {
       opts->allow_unauthenticated = true;
     } else if (arg_has_value(argv[i], "--iconv", &inline_value)) {
@@ -259,6 +261,28 @@ int server_cli_parse(int argc, char* argv[], ServerCliOptions* opts, char* err,
     set_error(err, err_size, "--hash-credentials cannot be combined with --daemon or --stdio");
     return -1;
   }
+  if (opts->allow_super && opts->no_super) {
+    set_error(err, err_size, "--allow-super and --no-super are mutually exclusive");
+    return -1;
+  }
+  if (opts->allow_super && opts->daemon_mode) {
+    set_error(err, err_size,
+              "--allow-super is for a locally-launched standalone TCP server; daemon modules opt "
+              "in per module with 'client owner = yes'");
+    return -1;
+  }
+  /* --stdio is the SSH transport: the remote server argv is composed by the
+   * CLIENT (directly and via --remote-option), so a client could otherwise pass
+   * --allow-super to a root --stdio receiver and defeat the C3 secure default.
+   * Never honor it there; the super mode stays forced OFF.  An operator who
+   * must keep the historical permissive behavior over SSH has to launch the
+   * receiver through a forced command, not via client-composed argv. */
+  if (opts->allow_super && opts->stdio_mode) {
+    set_error(err, err_size,
+              "--allow-super is not accepted with --stdio (the remote argv is client-composed; "
+              "use a forced command if the default must hold)");
+    return -1;
+  }
   if (opts->hash_iterations_set && opts->hash_credentials_file == NULL) {
     set_error(err, err_size, "--iterations requires --hash-credentials");
     return -1;

src/server/server_cli.h

@@ -45,6 +45,17 @@ typedef struct ServerCliOptions {
    * device-node creation) even when running as root.  Applies to --stdio and
    * --daemon alike; also makes the server refuse any client --copy-as. */
   bool no_super; /* --no-super */
+  /* --allow-super: locally-launched standalone TCP listener opt-in that keeps
+   * the historical permissive behavior for a PRIVILEGED (root) receiver.
+   * Without it a root standalone server forces SUPER_MODE_OFF, so a client
+   * --devices / --write-devices / --super / ownership request cannot make it
+   * create device nodes, write raw devices, or apply client-chosen ownership.
+   * It is rejected for --stdio: that path's remote argv is composed by the
+   * client (directly and via --remote-option), so it must never opt a root
+   * receiver back into super mode.  Non-root receivers are unaffected (the
+   * kernel refuses the confined attempts).  The daemon path instead uses the
+   * per-module `client owner = yes` opt-in. */
+  bool allow_super; /* --allow-super */
   /* --iconv=CONVERT_SPEC: the server's own LOCAL charset declaration.  The
    * client's full spec rides the wire config frame anyway; when the server is
    * started with its own --iconv, its LOCAL half overrides the local charset

src/shared/chunk.c

@@ -1,6 +1,7 @@
 #include <stddef.h>
 #include <stdint.h>
 #include <limits.h>
+#include <stdatomic.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -20,6 +21,33 @@
 #define MAX_FILE_DATA_SIZE (64ULL * 1024 * 1024)
 #define MAX_FILES_PER_CHUNK 65536U
 
+/* Reserve `charge` against `session`'s connection budget.  This mirrors the
+   static protocol_reserve_memory() in protocol.c: the receive-side call sites
+   only have the Data.owner pointer (a ProtocolSession*), and protocol.c is out
+   of scope for this fix, so the same atomic CAS accounting is reproduced here.
+   The matching release always goes through data_destroy()'s Data.owner path. */
+static bool chunk_session_reserve(ProtocolSession* session, size_t charge) {
+  unsigned long long allocated = atomic_load(&session->total_allocated_bytes);
+  while (true) {
+    if (allocated > MAX_CONNECTION_MEMORY ||
+        (unsigned long long)charge > MAX_CONNECTION_MEMORY - allocated)
+      return false;
+    if (atomic_compare_exchange_weak(&session->total_allocated_bytes, &allocated,
+                                     allocated + (unsigned long long)charge))
+      return true;
+  }
+}
+
+bool data_charge_session(Data* data, ProtocolSession* session, size_t charge) {
+  if (!data || charge == 0 || session == NULL)
+    return true;
+  if (!chunk_session_reserve(session, charge))
+    return false;
+  data->owner = session;
+  data->protocol_charge = charge;
+  return true;
+}
+
 Chunk* chunk_create(File** items, int element_count) {
   if (element_count < 0 || (element_count > 0 && items == NULL))
     return NULL;
@@ -370,6 +398,15 @@ Chunk* chunk_deserialize(Data* data, bool use_metadata) {
     Data* replacement = data_create(file_data, file_data_size);
     if (replacement == NULL)
       goto error;
+    /* Charge the retained per-file copy to the connection budget (when the
+       inbound chunk carries an owning session) so the queued copies are not
+       held outside MAX_CONNECTION_MEMORY (B6).  A NULL owner (e.g. a local
+       batch apply) leaves the copy uncharged. */
+    if (!data_charge_session(replacement, data->owner, allocation_size)) {
+      log_message(LOG_LEVEL_ERROR, "Per-connection memory limit exceeded for chunk file data");
+      data_destroy(replacement);
+      goto error;
+    }
     data_destroy(file->data);
     file->data = replacement;
     data_pointer += file_data_size;
@@ -466,12 +503,21 @@ Chunk* receive_chunk_data(int fd, const Config* config) {
   }
   Data* data_to_process = chunk_data;
   if (config->use_compression) {
+    /* Preserve the inbound session across decompression so the (larger)
+       decompressed chunk is charged to the same connection budget; the
+       compressed buffer's own charge is released by data_destroy below. */
+    ProtocolSession* owner = chunk_data->owner;
     data_to_process = data_decompress_limited(chunk_data, MAX_CHUNK_SIZE);
     data_destroy(chunk_data);
     if (data_to_process == NULL) {
       log_message(LOG_LEVEL_ERROR, "Failed to decompress chunk");
       return NULL;
     }
+    if (!data_charge_session(data_to_process, owner, data_to_process->size)) {
+      log_message(LOG_LEVEL_ERROR, "Per-connection memory limit exceeded for decompressed chunk");
+      data_destroy(data_to_process);
+      return NULL;
+    }
   }
 
   // Reject chunks larger than the maximum allowed size to prevent OOM.

src/shared/chunk.h

@@ -23,4 +23,15 @@ Data* chunk_compress_with_threads(Chunk* chunk, int compression_level, bool use_
                                   int compression_threads);
 Chunk* receive_chunk_data(int fd, const Config* config);
 
+/* Charge `charge` retained bytes of `data` against `session`'s per-connection
+ * budget (MAX_CONNECTION_MEMORY), mirroring the protocol layer's accounting, and
+ * record them on `data` so data_destroy() returns the charge through the
+ * Data.owner path.  Returns false (leaving `data` uncharged) when the ceiling
+ * would be exceeded.  A NULL/zero-size charge or a NULL session is a no-op
+ * success.  The receive-side decompression and chunk-copy paths know the owning
+ * session only through the Data.owner of the buffer they are processing, so
+ * this is the entry point that lets them participate in the connection budget
+ * without a session handle (B6). */
+bool data_charge_session(Data* data, ProtocolSession* session, size_t charge);
+
 #endif

src/shared/compression.c

@@ -243,9 +243,13 @@ Data* data_decompress_limited(Data* compressed_data, size_t maximum_size) {
   log_debug_message(LOG_DEBUG_UTIL, "Start to decompress data");
   unsigned long long dst_size =
       ZSTD_getFrameContentSize(compressed_data->data, compressed_data->size);
-  if (ZSTD_isError(dst_size)) {
-    log_message(LOG_LEVEL_ERROR, "Failed to get decompressed size: %s",
-                ZSTD_getErrorName(dst_size));
+  /* ZSTD_isError() is also true for ZSTD_CONTENTSIZE_ERROR and
+   * ZSTD_CONTENTSIZE_UNKNOWN (both are encoded near (size_t)-1), so test the
+   * sentinels explicitly instead of blanket-rejecting every error-ish value:
+   * only CONTENTSIZE_ERROR means an unreadable header, while CONTENTSIZE_UNKNOWN
+   * must reach the estimate fallback below. */
+  if (dst_size == ZSTD_CONTENTSIZE_ERROR) {
+    log_message(LOG_LEVEL_ERROR, "Failed to get decompressed size: invalid zstd frame");
     return NULL;
   }
 
@@ -324,6 +328,20 @@ Data* data_decompress_limited(Data* compressed_data, size_t maximum_size) {
       uncompressed_data->data = new_data;
       output.dst = new_data;
       output.size = buf_size;
+      /* Re-attempt with the larger output buffer; the truncated-frame check
+       * below must not reject a complete frame that merely filled the previous
+       * buffer exactly. */
+      continue;
+    }
+    /* A positive hint with all input consumed means the frame is incomplete: a
+     * truncated stream would otherwise spin here forever (ZSTD_decompressStream
+     * keeps returning the same hint).  Fail instead of burning CPU. */
+    if (ret != 0 && input.pos == input.size) {
+      log_message(LOG_LEVEL_ERROR,
+                  "Truncated zstd frame: input exhausted with %zu bytes still expected", ret);
+      data_destroy(uncompressed_data);
+      uncompressed_data = NULL;
+      goto cleanup;
     }
   } while (ret > 0);
 

src/shared/config.c

@@ -613,19 +613,36 @@ int config_parse_transport_dest(Config* config) {
   int daemon_ret = config_parse_daemon_dest(config);
   if (daemon_ret != 0)
     return daemon_ret;
-  config_parse_ssh_dest(config);
-  return 0;
+  /* 0 for a local destination (nothing parsed) or a valid SSH destination;
+   * -1 (already logged) for an injection-shaped user@host. */
+  return config_parse_ssh_dest(config);
 }
 
-void config_parse_ssh_dest(Config* config) {
+int config_parse_ssh_dest(Config* config) {
+  if (!config || !config->receive_root_directory)
+    return 0;
   if (!config_is_remote_dest(config->receive_root_directory))
-    return;
+    return 0;
+  const char* dest = config->receive_root_directory;
+  const char* colon = strchr(dest, ':');
+  /* The user@host token is passed to ssh in option position, so a user or host
+   * beginning with '-' would be consumed by ssh as an option (argument
+   * injection: e.g. "-oProxyCommand=...").  An empty host is likewise not a
+   * valid destination.  Validate before any wire/argv construction. */
+  const char* at = memchr(dest, '@', (size_t)(colon - dest));
+  const char* host = at ? at + 1 : dest;
+  size_t host_len = (size_t)(colon - host);
+  size_t user_len = at ? (size_t)(at - dest) : 0;
+  if (host_len == 0 || host[0] == '-' || (user_len > 0 && dest[0] == '-'))
+    return daemon_dest_parse_error("invalid remote destination user@host (must not be empty or "
+                                   "start with '-')",
+                                   dest);
   config->transport = TRANSPORT_SSH;
-  config->ssh_destination = str_dup(config->receive_root_directory);
-  const char* colon = strchr(config->receive_root_directory, ':');
+  config->ssh_destination = str_dup(dest);
   char* path = str_dup(colon + 1);
   free(config->receive_root_directory);
   config->receive_root_directory = path;
+  return 0;
 }
 
 void config_burn_auth(Config* config) {

src/shared/config.h

@@ -851,7 +851,10 @@ bool config_send(int file_descriptor, const Config* config);
 bool config_send_wire_block(int file_descriptor, const Config* config);
 Config* config_receive(int file_descriptor);
 bool config_is_remote_dest(const char* s);
-void config_parse_ssh_dest(Config* config);
+/* Parse a single-colon host:path SSH destination (0 = not an SSH destination or
+ * parsed successfully, -1 = rejected, e.g. a user@host beginning with '-'; the
+ * reason is logged). */
+int config_parse_ssh_dest(Config* config);
 
 /* A ConfigValidateFunc may return this sentinel to tell
  * config_receive_with_validate that the callback ALREADY sent a terminal status

src/shared/credentials.c

@@ -158,13 +158,13 @@ static int hex_value(char c) {
  * digits.  Such a line is refused loudly (and never accepted) so an operator
  * cannot keep a replayable bearer digest in place after the protocol bump. */
 static bool secret_is_legacy_hex(const char* s) {
-  if (!s)
+  if (!s || strlen(s) != 64)
     return false;
   for (int i = 0; i < 64; i++) {
     if (hex_value(s[i]) < 0)
       return false;
   }
-  return s[64] == '\0';
+  return true;
 }
 
 bool credentials_b64_encode(const uint8_t* in, size_t n, char* out, size_t out_sz) {

src/shared/daemon_conf.c

@@ -133,6 +133,7 @@ static bool store_host_list(char*** list, int* count, const char* value, const c
     return false;
   }
   char* save = NULL;
+  int added = 0;
   for (char* token = strtok_r(copy, ", \t", &save); token; token = strtok_r(NULL, ", \t", &save)) {
     if (!host_pattern_valid(token)) {
       if (module_name)
@@ -163,8 +164,20 @@ static bool store_host_list(char*** list, int* count, const char* value, const c
       return false;
     }
     (*list)[(*count)++] = dup;
+    added++;
   }
   free(copy);
+  /* A present key with an empty (or separator-only) value would otherwise
+   * install a zero-length list, i.e. no ACL at all: a strict-parse config must
+   * never silently turn a restrictive directive into "allow everyone". */
+  if (added == 0) {
+    if (module_name)
+      set_error(err, err_size, "module '%s': '%s' must list at least one host pattern", module_name,
+                key);
+    else
+      set_error(err, err_size, "'%s' must list at least one host pattern", key);
+    return false;
+  }
   return true;
 }
 
@@ -403,6 +416,7 @@ static bool apply_module_key(DaemonModule* module, char* key, char* value, char*
       return false;
     }
     char* save = NULL;
+    int added = 0;
     for (char* token = strtok_r(list, ",", &save); token; token = strtok_r(NULL, ",", &save)) {
       const char* user = trim_ws(token);
       if (*user == '\0')
@@ -430,8 +444,16 @@ static bool apply_module_key(DaemonModule* module, char* key, char* value, char*
         return false;
       }
       module->auth_users[module->auth_user_count++] = dup;
+      added++;
     }
     free(list);
+    /* An empty/separator-only value must not silently disable authentication:
+     * the key's presence is an explicit request for an allow-list. */
+    if (added == 0) {
+      set_error(err, err_size, "module '%s': 'auth users' must list at least one user",
+                module->name);
+      return false;
+    }
     return true;
   }
   if (key_equals(key, "max connections"))
@@ -439,10 +461,10 @@ static bool apply_module_key(DaemonModule* module, char* key, char* value, char*
                               "max connections", module->name, err, err_size);
   if (key_equals(key, "hosts allow"))
     return store_host_list(&module->hosts_allow, &module->hosts_allow_count, value, "hosts allow",
-                           false, module->name, err, err_size);
+                           module->name, false, err, err_size);
   if (key_equals(key, "hosts deny"))
     return store_host_list(&module->hosts_deny, &module->hosts_deny_count, value, "hosts deny",
-                           false, module->name, err, err_size);
+                           module->name, false, err, err_size);
   set_error(err, err_size, "unknown key '%s' in module '%s'", key, module->name);
   return false;
 }

src/shared/file.c

@@ -890,13 +890,35 @@ static bool file_to_disk_secure_impl(const char* path, const void* data,
   if (inplace) {
     /* --inplace writes directly into the destination; a scratch --temp-dir
        does not apply and must never redirect these writes. */
-    fd = openat(dirfd, leaf, O_WRONLY | O_CREAT | O_CLOEXEC | O_NOFOLLOW, 0644);
+    /* Type gate BEFORE opening: an existing destination entry that is not a
+       regular file (FIFO, socket, char/block device, directory) must never be
+       opened for writing.  Opening a FIFO would block the receive thread
+       forever and writing into a device would bypass the --write-devices /
+       super-mode gate (a client-controlled device write).  fstatat with
+       AT_SYMLINK_NOFOLLOW does not follow a symlink and does not block. */
+    struct stat pre_stat;
+    if (fstatat(dirfd, leaf, &pre_stat, AT_SYMLINK_NOFOLLOW) == 0 && !S_ISREG(pre_stat.st_mode)) {
+      close(dirfd);
+      free(leaf);
+      return false;
+    }
+    /* O_NONBLOCK: a no-op for a regular file, but a raced-in FIFO cannot block
+       the open before the post-open S_ISREG re-check rejects it. */
+    fd = openat(dirfd, leaf, O_WRONLY | O_CREAT | O_CLOEXEC | O_NOFOLLOW | O_NONBLOCK, 0644);
     if (fd >= 0) {
       struct stat destination_stat;
+      /* Re-check the opened descriptor: a concurrent replacement between the
+         fstatat probe and the open (or a device/FIFO raced in) must never be
+         written through. */
+      if (fstat(fd, &destination_stat) != 0 || !S_ISREG(destination_stat.st_mode)) {
+        close(fd);
+        close(dirfd);
+        free(leaf);
+        return false;
+      }
       bool newer = false;
-      if (update && metadata && fstat(fd, &destination_stat) == 0 &&
-          S_ISREG(destination_stat.st_mode)) {
-        newer = stat_is_newer(&destination_stat, metadata);
+      if (update && metadata && stat_is_newer(&destination_stat, metadata)) {
+        newer = true;
       }
       if (newer) {
         ok = true;
@@ -1226,11 +1248,19 @@ static bool file_to_disk_secure_link_impl(const char* path, const char* basis_pa
       if (linked) {
         int target_dirfd = scratch_dirfd >= 0 ? scratch_dirfd : dirfd;
         if (use_fsync) {
-          int tfd = openat(target_dirfd, tmp, O_RDONLY | O_NOFOLLOW | O_CLOEXEC);
-          if (tfd < 0 || fsync(tfd) != 0) {
+          /* O_NONBLOCK: the freshly linked temp is normally the basis's regular
+             file, but a raced-in FIFO at the name must not block this reopen
+             forever.  With O_NONBLOCK such an open fails with ENXIO instead of
+             blocking, which is treated as a benign fsync-skip (the link itself
+             is still installed); any other open/fsync failure falls back to the
+             byte-copy path as before. */
+          int tfd = openat(target_dirfd, tmp, O_RDONLY | O_NOFOLLOW | O_CLOEXEC | O_NONBLOCK);
+          if (tfd < 0) {
+            if (errno != ENXIO)
+              linked = false;
+          } else if (fsync(tfd) != 0) {
             linked = false;
-            if (tfd >= 0)
-              close(tfd);
+            close(tfd);
           } else {
             close(tfd);
           }

src/shared/file_list.c

@@ -23,6 +23,8 @@ static void string_list_destroy(StringList* list) {
 
 static bool string_list_add(StringList* list, const char* text) {
   if (list->count == list->capacity) {
+    if (list->capacity > INT_MAX / 2)
+      return false;
     int new_cap = list->capacity > 0 ? list->capacity * 2 : 16;
     char** grown = realloc(list->items, (size_t)new_cap * sizeof(char*));
     if (!grown)
@@ -56,8 +58,14 @@ static int normalize_entry(const char* raw, size_t len, bool strip_line_endings,
     snprintf(err, err_size, "absolute path entries are not allowed: '%.*s'", print_len, raw);
     return -1;
   }
-  /* Reject NUL bytes inside a token defensively (NUL-delimited mode splits on
-   * them, so this only guards against embedded garbage). */
+  /* Reject NUL bytes inside a token defensively.  In NUL-delimited mode the
+   * delimiter itself is the final byte and is expected; in line mode any NUL is
+   * embedded garbage (strlen-based parsing would otherwise silently truncate). */
+  size_t scan_len = strip_line_endings ? len : len - 1;
+  if (memchr(raw, '\0', scan_len)) {
+    snprintf(err, err_size, "entry contains an embedded NUL byte");
+    return -1;
+  }
   char* dup = malloc(len + 1);
   if (!dup) {
     snprintf(err, err_size, "memory allocation failed");
@@ -158,10 +166,20 @@ FileListSet* file_list_load(const char* path, bool null_separated, char* err, si
   StringList raw = {0};
   char* line = NULL;
   size_t line_cap = 0;
-  ssize_t n;
   bool ok = true;
   char delim = null_separated ? '\0' : '\n';
-  while (ok && (n = getdelim(&line, &line_cap, delim, fp)) != -1) {
+  while (ok) {
+    ssize_t n = utils_getdelim_bounded(fp, &line, &line_cap, delim, UTILS_MAX_LINE_LEN);
+    if (n < 0) {
+      if (errno == EFBIG)
+        snprintf(err, err_size, "entry in file list exceeds %d bytes", (int)UTILS_MAX_LINE_LEN);
+      else
+        snprintf(err, err_size, "error reading file list: %s", strerror(errno));
+      ok = false;
+      break;
+    }
+    if (n == 0)
+      break;
     int r = normalize_entry(line, (size_t)n, !null_separated, &raw, err, err_size);
     if (r < 0) {
       ok = false;

src/shared/file_receive.c

@@ -12,6 +12,7 @@
 #include "array_list.h"
 #include "charset.h"
 #include "chmod.h"
+#include "chunk.h"
 #include "compression.h"
 #include "config.h"
 #include "data.h"
@@ -27,6 +28,11 @@
 
 #define MAX_SERVER_DELETE_COUNT 100000U
 #define MAX_FILE_DATA_SIZE MAX_RECEIVE_WHOLE_FILE_SIZE
+/* Retained cost of one delete-manifest entry beyond its path bytes: the
+   ArrayList pointer slot plus an approximate malloc header/rounding for the
+   heap copy.  Charged against MAX_MANIFEST_BYTES so a frame full of tiny paths
+   cannot retain far more than the byte budget (B5). */
+#define MANIFEST_ENTRY_OVERHEAD (sizeof(char*) + 16)
 
 bool file_save_to_disk(const char* root_directory, const File* file, const Config* config) {
   return file_save_to_disk_full(root_directory, file, config) != FILE_SAVE_ERROR;
@@ -127,7 +133,10 @@ static bool hardlink_read_source(const char* path, void** out_buf, unsigned long
     *source_absent = errno == ENOENT || errno == ENOTDIR;
     return false;
   }
-  int fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW);
+  /* O_NONBLOCK is a no-op for a regular file but makes openat() fail/succeed
+     immediately for a client-planted FIFO instead of blocking the receive
+     thread forever; the post-open S_ISREG gate below is the actual type check. */
+  int fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW | O_NONBLOCK);
   int saved_errno = errno;
   free(leaf);
   close(parent_fd);
@@ -254,7 +263,8 @@ static FileSaveResult file_save_hardlink_sibling(const char* root_directory, con
       free(destination_path);
       return absent_result;
     }
-    FileXattrList* sibling_xattrs = cfg->use_xattrs ? xattr_capture_path(staged_first) : NULL;
+    FileXattrList* sibling_xattrs =
+        cfg->use_xattrs ? xattr_capture_path(staged_first, cfg->preserve_acls) : NULL;
     bool ok = file_to_disk_secure_link_attrs(
         staged_sibling, staged_first, content, content_size, preallocate, file->metadata,
         preserve_executability, use_fsync, sibling_xattrs, cfg ? cfg->fake_super : false, NULL);
@@ -286,7 +296,8 @@ static FileSaveResult file_save_hardlink_sibling(const char* root_directory, con
     return absent_result;
   }
   const char* temp_dir = (cfg && cfg->temp_dir) ? cfg->temp_dir : NULL;
-  FileXattrList* sibling_xattrs = cfg->use_xattrs ? xattr_capture_path(first_disk) : NULL;
+  FileXattrList* sibling_xattrs =
+      cfg->use_xattrs ? xattr_capture_path(first_disk, cfg->preserve_acls) : NULL;
   bool ok = file_to_disk_secure_link_attrs(
       destination_path, first_disk, content, content_size, preallocate, file->metadata,
       preserve_executability, use_fsync, sibling_xattrs, cfg ? cfg->fake_super : false, temp_dir);
@@ -944,7 +955,7 @@ static bool receive_file_xattrs(File* file, int fd, const Config* config) {
   if (!config->use_xattrs)
     return true;
   int xok = 0;
-  FileXattrList* list = xattr_receive(fd, &xok);
+  FileXattrList* list = xattr_receive(fd, &xok, config->preserve_acls);
   if (!xok) {
     xattr_list_free(list);
     return false;
@@ -1009,6 +1020,7 @@ static File* receive_delta_file(int fd, const Config* config, const char* check_
         !compression_should_skip_with_suffixes(
             check_path, config->skip_compress_suffixes,
             config->skip_compress_set ? config->skip_compress_count : -1)) {
+      ProtocolSession* owner = delta_data->owner;
       raw_delta = data_decompress_limited(delta_data, MAX_RECEIVE_WHOLE_FILE_SIZE);
       data_destroy(delta_data);
       if (!raw_delta) {
@@ -1017,6 +1029,15 @@ static File* receive_delta_file(int fd, const Config* config, const char* check_
         *failed = true;
         return NULL;
       }
+      /* Charge the decompressed delta to the connection budget (the paired
+         wire buffer's charge was just released). */
+      if (!data_charge_session(raw_delta, owner, raw_delta->size)) {
+        data_destroy(raw_delta);
+        free(old_data);
+        delta_signature_destroy(sig);
+        *failed = true;
+        return NULL;
+      }
     }
 
     Delta* delta = delta_deserialize(raw_delta);
@@ -1131,12 +1152,20 @@ static File* receive_delta_file(int fd, const Config* config, const char* check_
             file->path, config->skip_compress_suffixes,
             config->skip_compress_set ? config->skip_compress_count : -1)) {
       Data* uncompressed = data_decompress_limited(file_data, MAX_RECEIVE_WHOLE_FILE_SIZE);
+      ProtocolSession* owner = file_data->owner;
       data_destroy(file_data);
       if (uncompressed == NULL) {
         file_destroy(file);
         *failed = true;
         return NULL;
       }
+      if (!data_charge_session(uncompressed, owner, uncompressed->size)) {
+        data_destroy(uncompressed);
+        file_destroy(file);
+        send_status(fd, STATUS_ERROR);
+        *failed = true;
+        return NULL;
+      }
       if (uncompressed->size > MAX_FILE_DATA_SIZE) {
         data_destroy(uncompressed);
         file_destroy(file);
@@ -1193,7 +1222,9 @@ static bool basis_open_regular(const char* path, unsigned long long expected_siz
   int parent_fd = file_open_secure_parent(path, &leaf, false);
   if (parent_fd < 0)
     return false;
-  int fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW);
+  /* O_NONBLOCK: a client-planted FIFO must not block the receiver's openat()
+     forever; the fstat()/S_ISREG gate below rejects it immediately. */
+  int fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW | O_NONBLOCK);
   free(leaf);
   close(parent_fd);
   if (fd < 0)
@@ -1247,14 +1278,27 @@ static bool basis_quick_matches(const Config* config, const struct stat* st, tim
 
 /* Search the basis-dir list in command-line order and return the first exact
    match.  When load_content is true the matched bytes are kept in out->content
-   so the caller can materialize the file without re-reading it. */
+   so the caller can materialize the file without re-reading it.
+
+   An exact match ALSO requires the basis bytes' digest to equal the source's,
+   so `hash_content` gates the content read/hash itself.  A server-contacting
+   --dry-run passes hash_content=false: no basis file may be read or hashed
+   (that would be a 1-bit content oracle against a client-supplied digest), so a
+   metadata-only pass can never confirm a hit and declines it.  The real path
+   always passes hash_content=true, keeping its behavior byte-for-byte. */
 static bool basis_match_find(const Config* config, const char* check_path,
                              unsigned long long check_size, time_t check_mtime,
                              long check_mtime_nsec, const uint8_t* check_digest,
-                             size_t check_digest_len, bool load_content, BasisMatch* out) {
+                             size_t check_digest_len, bool load_content, bool hash_content,
+                             BasisMatch* out) {
   memset(out, 0, sizeof(*out));
   if (!config || !config_has_basis(config) || config->ignore_times)
     return false;
+  /* Dry-run: never read/hash basis content.  A hit cannot be decided from
+     metadata alone, so report no match (the caller treats it as would-transfer)
+     without touching the file's contents. */
+  if (!hash_content)
+    return false;
   for (int i = 0; i < config->basis_count; i++) {
     const BasisDest* entry = &config->basis_dirs[i];
     char* basis_dir = path_cat(config->receive_root_directory, entry->path);
@@ -1637,11 +1681,17 @@ static File* receive_full_file(int fd, const Config* config, const char* path) {
                                              config->skip_compress_set ? config->skip_compress_count
                                                                        : -1)) {
     Data* uncompressed = data_decompress_limited(file_data, MAX_RECEIVE_WHOLE_FILE_SIZE);
+    ProtocolSession* owner = file_data->owner;
     data_destroy(file_data);
     if (uncompressed == NULL) {
       file_destroy(file);
       return NULL;
     }
+    if (!data_charge_session(uncompressed, owner, uncompressed->size)) {
+      data_destroy(uncompressed);
+      file_destroy(file);
+      return NULL;
+    }
     if (uncompressed->size > MAX_FILE_DATA_SIZE) {
       data_destroy(uncompressed);
       file_destroy(file);
@@ -1776,7 +1826,9 @@ static IncrementalCheckOutcome incremental_check_open_destination(IncrementalChe
   char* leaf = NULL;
   int parent_fd = file_open_secure_parent(full_path, &leaf, false);
   if (parent_fd >= 0) {
-    state->old_fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW);
+    /* O_NONBLOCK: an existing FIFO at the destination must not block this
+       openat(); the S_ISREG gate below rejects the non-regular entry. */
+    state->old_fd = openat(parent_fd, leaf, O_RDONLY | O_CLOEXEC | O_NOFOLLOW | O_NONBLOCK);
     free(leaf);
     close(parent_fd);
     state->has_old_file = state->old_fd >= 0 && fstat(state->old_fd, &state->old_st) == 0 &&
@@ -1815,7 +1867,15 @@ static IncrementalCheckOutcome incremental_check_quick_skip(IncrementalCheckStat
   bool try_delta = config->use_delta && !config->whole_file && has_old_file &&
                    delta_should_attempt(old_size, state->check_size, config->delta_max_file_size);
   bool checksum_needs_read = size_equal && !config->ignore_times && config->checksum;
-  bool need_old_data = checksum_needs_read || try_delta;
+  /* --dry-run must never read the destination file's CONTENTS: a client could
+     otherwise use `--dry-run --checksum` against a read-only module as a
+     1-bit content oracle (hash match / mismatch) and force arbitrary reads.
+     Decide from metadata alone; when metadata is inconclusive (checksum or
+     delta would have required the body) report would-transfer.  The real
+     (non-dry-run) behavior below is unchanged. */
+  bool need_old_data = !config->dry_run && (checksum_needs_read || try_delta);
+  if (config->dry_run)
+    try_delta = false;
   *out_try_delta = try_delta;
 
   if (need_old_data && has_old_file && old_size > 0 && old_size <= MAX_RECEIVE_WHOLE_FILE_SIZE &&
@@ -1836,7 +1896,12 @@ static IncrementalCheckOutcome incremental_check_quick_skip(IncrementalCheckStat
   }
 
   bool match = false;
-  if (checksum_needs_read) {
+  if (config->dry_run) {
+    /* Metadata-only decision: a size match plus a matching mtime is treated as
+       up to date; --checksum/--delta cannot be verified without reading, so an
+       otherwise inconclusive comparison is a would-transfer. */
+    match = size_equal && !config->ignore_times && (config->size_only || match_by_metadata);
+  } else if (checksum_needs_read) {
     uint8_t old_digest[CHECKSUM_MAX_DIGEST_LEN];
     size_t old_len = 0;
     bool hashed = checksum_digest((ChecksumAlgo)config->checksum_algo, config->checksum_seed,
@@ -1861,10 +1926,13 @@ static IncrementalCheckOutcome incremental_check_quick_skip(IncrementalCheckStat
    When dry_run is set and the file is not already up to date the receiver must
    materialize nothing (no basis link/copy, no append/delta/full transfer) and
    the sender must send no data, so answer STATUS_DRY_RUN_TRANSFER and stop.
-   The one exception is a --compare-dest exact hit with no destination copy: a
-   real run would suppress the data without changing the destination, so it
-   reports as a skip (STATUS_OK) exactly as the full basis path below would.
-   Everything read here (destination file, basis candidates) is read-only. */
+
+   The basis lookup is deliberately content-blind: a real run would only accept
+   a --compare-dest exact hit after hashing the basis file and comparing it with
+   the client-supplied digest, which in a dry-run is a 1-bit content oracle.
+   Under dry_run no basis bytes may be read, so an otherwise-matching entry is
+   treated as would-transfer instead of a skip.  Everything read here (the
+   destination file's metadata, basis candidates' metadata) is read-only. */
 static IncrementalCheckOutcome incremental_check_dry_run_shortcut(IncrementalCheckState* state,
                                                                   bool* skipped,
                                                                   bool* would_transfer) {
@@ -1875,9 +1943,12 @@ static IncrementalCheckOutcome incremental_check_dry_run_shortcut(IncrementalChe
   bool skip_via_compare = false;
   if (config_has_basis(config) && !config->ignore_times) {
     BasisMatch basis;
+    /* hash_content=false: a dry-run must not read or hash the basis file.  No
+       content comparison is possible, so no compare-dest hit can be confirmed
+       and an otherwise-matching file is reported as would-transfer. */
     basis_match_find(config, state->check_path, state->check_size, (time_t)state->check_mtime,
                      (long)state->check_mtime_nsec, state->check_digest, state->check_digest_len,
-                     false, &basis);
+                     false, false, &basis);
     if (basis.hit && basis.type == BASIS_DEST_COMPARE && !state->has_old_file)
       skip_via_compare = true;
     basis_match_free(&basis);
@@ -1905,7 +1976,7 @@ static IncrementalCheckOutcome incremental_check_try_basis(IncrementalCheckState
   BasisMatch basis;
   basis_match_find(config, state->check_path, state->check_size, (time_t)state->check_mtime,
                    (long)state->check_mtime_nsec, state->check_digest, state->check_digest_len,
-                   true, &basis);
+                   true, true, &basis);
   if (basis.hit) {
     if (basis.type == BASIS_DEST_COMPARE) {
       basis_match_free(&basis);
@@ -2050,7 +2121,7 @@ static IncrementalCheckOutcome incremental_check_try_append_resume(IncrementalCh
   }
   if (config->use_xattrs) {
     int xok = 0;
-    append_xattrs = xattr_receive(fd, &xok);
+    append_xattrs = xattr_receive(fd, &xok, config->preserve_acls);
     if (!xok) {
       xattr_list_free(append_xattrs);
       return INCREMENTAL_ERROR;
@@ -2066,11 +2137,17 @@ static IncrementalCheckOutcome incremental_check_try_append_resume(IncrementalCh
                                              config->skip_compress_set ? config->skip_compress_count
                                                                        : -1)) {
     Data* uncompressed = data_decompress_limited(tail, MAX_RECEIVE_WHOLE_FILE_SIZE);
+    ProtocolSession* owner = tail->owner;
     data_destroy(tail);
     if (uncompressed == NULL) {
       xattr_list_free(append_xattrs);
       return INCREMENTAL_ERROR;
     }
+    if (!data_charge_session(uncompressed, owner, uncompressed->size)) {
+      data_destroy(uncompressed);
+      xattr_list_free(append_xattrs);
+      return INCREMENTAL_ERROR;
+    }
     if (uncompressed->size > MAX_FILE_DATA_SIZE) {
       data_destroy(uncompressed);
       xattr_list_free(append_xattrs);
@@ -2301,11 +2378,17 @@ File* file_receive(const Config* config, int file_descriptor) {
                                              config->skip_compress_set ? config->skip_compress_count
                                                                        : -1)) {
     Data* file_data_uncompressed = data_decompress_limited(file_data, MAX_RECEIVE_WHOLE_FILE_SIZE);
+    ProtocolSession* owner = file_data->owner;
     data_destroy(file_data);
     if (file_data_uncompressed == NULL) {
       file_destroy(file);
       return NULL;
     }
+    if (!data_charge_session(file_data_uncompressed, owner, file_data_uncompressed->size)) {
+      data_destroy(file_data_uncompressed);
+      file_destroy(file);
+      return NULL;
+    }
     if (file_data_uncompressed->size > MAX_FILE_DATA_SIZE) {
       data_destroy(file_data_uncompressed);
       file_destroy(file);
@@ -2694,19 +2777,21 @@ File* file_receive_special(int file_descriptor) {
    manifest).  Returns an owned DeleteManifest, or NULL after sending STATUS_ERROR
    when the frame is malformed (bad count, empty/absolute path, path traversal,
    or an aggregate size beyond MAX_MANIFEST_BYTES). */
-static bool receive_manifest_section(int fd, ArrayList* list, size_t* manifest_bytes) {
+static bool receive_manifest_section(int fd, ArrayList* list, size_t* manifest_bytes,
+                                     size_t* manifest_entries) {
   int count;
   if (!receive_int(fd, &count)) {
     send_status(fd, STATUS_ERROR);
     return false;
   }
-  if (count < 0 || count > MAX_MANIFEST_ENTRIES) {
+  if (count < 0 || count > MAX_MANIFEST_ENTRIES ||
+      (size_t)count > MAX_MANIFEST_ENTRIES - *manifest_entries) {
     send_status(fd, STATUS_ERROR);
     return false;
   }
   for (int i = 0; i < count; i++) {
     char* s = receive_wire_str(fd);
-    size_t entry_size = s ? strlen(s) : 0;
+    size_t entry_size = s ? strlen(s) + MANIFEST_ENTRY_OVERHEAD : 0;
     if (!s || s[0] == '\0' || s[0] == '/' || has_path_traversal(s) ||
         entry_size > MAX_MANIFEST_BYTES - *manifest_bytes ||
         (*manifest_bytes += entry_size) > MAX_MANIFEST_BYTES || !array_list_add(list, s)) {
@@ -2715,6 +2800,7 @@ static bool receive_manifest_section(int fd, ArrayList* list, size_t* manifest_b
       return false;
     }
   }
+  *manifest_entries += (size_t)count;
   return true;
 }
 
@@ -2733,9 +2819,10 @@ DeleteManifest* receive_manifest_entries(int fd) {
     return NULL;
   }
   size_t manifest_bytes = 0;
-  if (!receive_manifest_section(fd, manifest->keeps, &manifest_bytes) ||
-      !receive_manifest_section(fd, manifest->protected, &manifest_bytes) ||
-      !receive_manifest_section(fd, manifest->missing, &manifest_bytes)) {
+  size_t manifest_entries = 0;
+  if (!receive_manifest_section(fd, manifest->keeps, &manifest_bytes, &manifest_entries) ||
+      !receive_manifest_section(fd, manifest->protected, &manifest_bytes, &manifest_entries) ||
+      !receive_manifest_section(fd, manifest->missing, &manifest_bytes, &manifest_entries)) {
     delete_manifest_free(manifest);
     return NULL;
   }

src/shared/filter.c

@@ -2,6 +2,7 @@
 #include "log.h"
 #include "utils.h"
 #include <errno.h>
+#include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -176,6 +177,8 @@ bool filter_rule_list_add(FilterRuleList* list, FilterRule* rule) {
   if (!list || !rule)
     return false;
   if (list->count == list->capacity) {
+    if (list->capacity > INT_MAX / 2)
+      return false;
     int new_cap = list->capacity > 0 ? list->capacity * 2 : 8;
     FilterRule** grown = realloc(list->items, (size_t)new_cap * sizeof(FilterRule*));
     if (!grown)
@@ -322,8 +325,10 @@ FilterRuleList* filter_file_read(const char* dir_path, const char* owner_rel, bo
   if (!fp) {
     if (errno == ENOENT || errno == ENOTDIR)
       return filter_rule_list_create();
-    log_message(LOG_LEVEL_WARNING, "Could not read .rsync-filter in %s: %s", dir_path,
-                strerror(errno));
+    char* escaped_dir = output_escape(dir_path, log_get_8_bit_output());
+    log_message(LOG_LEVEL_WARNING, "Could not read .rsync-filter in %s: %s",
+                escaped_dir ? escaped_dir : "<allocation failed>", strerror(errno));
+    free(escaped_dir);
     return filter_rule_list_create();
   }
   if (exists)
@@ -336,9 +341,20 @@ FilterRuleList* filter_file_read(const char* dir_path, const char* owner_rel, bo
   }
   char* line = NULL;
   size_t line_cap = 0;
-  ssize_t n;
   bool ok = true;
-  while ((n = getline(&line, &line_cap, fp)) != -1) {
+  while (true) {
+    ssize_t n = utils_getdelim_bounded(fp, &line, &line_cap, '\n', UTILS_MAX_LINE_LEN);
+    if (n < 0) {
+      if (errno == EFBIG) {
+        snprintf(err, err_size, "line in .rsync-filter exceeds %d bytes", (int)UTILS_MAX_LINE_LEN);
+      } else {
+        snprintf(err, err_size, "error reading .rsync-filter: %s", strerror(errno));
+      }
+      ok = false;
+      break;
+    }
+    if (n == 0)
+      break;
     const char* p = line;
     while (*p == ' ' || *p == '\t')
       p++;

src/shared/protocol.c

@@ -302,10 +302,14 @@ bool protocol_send_n_data(ProtocolSession* session, const void* data, size_t dat
     if (pfd.revents & (POLLERR | POLLNVAL))
       return false;
     ssize_t bytes_send;
-    if (session->ssl)
-      bytes_send = SSL_write(session->ssl, (const char*)data + total_bytes_send, chunk);
-    else
+    if (session->ssl) {
+      /* SSL_write takes an int length; clamp a >INT_MAX request into chunks so
+       * the size_t downcast can never truncate into a negative/partial write. */
+      size_t ssl_chunk = chunk > (size_t)INT_MAX ? (size_t)INT_MAX : chunk;
+      bytes_send = SSL_write(session->ssl, (const char*)data + total_bytes_send, (int)ssl_chunk);
+    } else {
       bytes_send = write(fd, (const char*)data + total_bytes_send, chunk);
+    }
     if (bytes_send <= 0) {
       if (session->ssl) {
         int ssl_err = SSL_get_error(session->ssl, (int)bytes_send);
@@ -387,6 +391,13 @@ static bool protocol_receive_n_data_until(ProtocolSession* session, void* data,
           wait_events = ssl_err == SSL_ERROR_WANT_WRITE ? POLLOUT : POLLIN;
           continue;
         }
+        /* A signal interrupts the blocking TLS read: retry (mirrors the send
+           path and protocol_read_status_until) so the loop reaches its next
+           abort/deadline checkpoint instead of failing spuriously. */
+        if (ssl_err == SSL_ERROR_SYSCALL && errno == EINTR)
+          continue;
+      } else if (errno == EINTR) {
+        continue;
       }
       if (bytes_received == 0)
         log_message(LOG_LEVEL_ERROR, "Connection closed while receiving data");

src/shared/transport_ssh.c

@@ -75,6 +75,15 @@ static int parse_remote_dest(const char* dest, RemoteDest* r) {
     memcpy(r->host, dest, host_len);
     r->host[host_len] = '\0';
   }
+  /* The user@host token is handed to ssh in option position.  Reject anything
+   * that ssh would consume as an option (a leading '-') or an empty host, so a
+   * crafted destination can never inject an ssh option such as
+   * -oProxyCommand=... .  This mirrors config_parse_ssh_dest's validation and
+   * is defense-in-depth for callers that bypass it. */
+  if (r->host[0] == '\0' || r->host[0] == '-' || r->user[0] == '-') {
+    remote_dest_destroy(r);
+    return -1;
+  }
   return 0;
 }
 
@@ -216,10 +225,10 @@ char** ssh_build_client_argv(const char* rsh_command, int port, const char* user
     nwords = 1;
   }
 
-  /* Fixed tail: three -o pairs (6) + optional -p/value (2) + user@host +
-   * remote command + terminating NULL. */
+  /* Fixed tail: three -o pairs (6) + optional -p/value (2) + the "--" end of
+   * options marker + user@host + remote command + terminating NULL. */
   int port_extra = (port > 0 && port != 22) ? 2 : 0;
-  size_t total = (size_t)nwords + 6 + (size_t)port_extra + 3;
+  size_t total = (size_t)nwords + 6 + (size_t)port_extra + 4;
   char** argv = calloc(total, sizeof(char*));
   if (!argv) {
     for (int i = 0; i < nwords; i++)
@@ -253,6 +262,13 @@ char** ssh_build_client_argv(const char* rsh_command, int port, const char* user
       goto fail_argv;
     ac++;
   }
+  /* End of options: guarantees the user@host token that follows is treated as
+   * the destination and never re-interpreted as an ssh option, even if every
+   * caller-side validation were bypassed. */
+  argv[ac] = str_dup("--");
+  if (!argv[ac])
+    goto fail_argv;
+  ac++;
   argv[ac] = str_dup(userhost);
   if (!argv[ac])
     goto fail_argv;

src/shared/transport_tls.c

@@ -4,7 +4,9 @@
 #include "transport_tcp.h"
 #include "utils.h"
 #include <arpa/inet.h>
+#include <fcntl.h>
 #include <openssl/err.h>
+#include <openssl/pem.h>
 #include <openssl/ssl.h>
 #include <signal.h>
 #include <stdio.h>
@@ -34,6 +36,59 @@ static void log_ssl_errors(void) {
   }
 }
 
+/* Load the TLS private key through an already-opened, no-follow descriptor so
+ * the owner/mode policy is checked on the SAME file object that is loaded: an
+ * attacker cannot swap the path between a stat() and a later open() (TOCTOU).
+ * The exact-owner / 0600 policy is preserved and group/other execute bits are
+ * rejected as well.  Ownership of the descriptor passes to the BIO and is
+ * released exactly once by BIO_free() (BIO_CLOSE). */
+static bool load_private_key_secure(SSL_CTX* ctx, const char* key) {
+  int fd = open(key, O_RDONLY | O_NOFOLLOW | O_CLOEXEC);
+  if (fd < 0) {
+    char* escaped = output_escape(key, false);
+    log_message(LOG_LEVEL_ERROR, "Failed to open private key: %s",
+                escaped ? escaped : "<allocation failed>");
+    free(escaped);
+    return false;
+  }
+  struct stat key_stat;
+  if (fstat(fd, &key_stat) != 0 || !S_ISREG(key_stat.st_mode) || key_stat.st_uid != geteuid() ||
+      (key_stat.st_mode & (S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH | S_IXGRP | S_IXOTH))) {
+    log_message(LOG_LEVEL_ERROR,
+                "TLS private key must be a regular file owned by the current user and private "
+                "(mode 0600)");
+    close(fd);
+    return false;
+  }
+  BIO* bio = BIO_new_fd(fd, BIO_CLOSE);
+  if (!bio) {
+    close(fd);
+    log_message(LOG_LEVEL_ERROR, "Failed to read private key");
+    return false;
+  }
+  EVP_PKEY* pkey = PEM_read_bio_PrivateKey(bio, NULL, NULL, NULL);
+  BIO_free(bio); /* releases fd via BIO_CLOSE */
+  if (!pkey) {
+    char* escaped = output_escape(key, false);
+    log_message(LOG_LEVEL_ERROR, "Failed to load private key: %s",
+                escaped ? escaped : "<allocation failed>");
+    free(escaped);
+    log_ssl_errors();
+    return false;
+  }
+  int use_ok = SSL_CTX_use_PrivateKey(ctx, pkey);
+  EVP_PKEY_free(pkey);
+  if (use_ok != 1) {
+    char* escaped = output_escape(key, false);
+    log_message(LOG_LEVEL_ERROR, "Failed to use private key: %s",
+                escaped ? escaped : "<allocation failed>");
+    free(escaped);
+    log_ssl_errors();
+    return false;
+  }
+  return true;
+}
+
 static SSL_CTX* create_ssl_ctx(bool is_server, const char* cert, const char* key,
                                const char* ca_path) {
   if (!is_server && !ca_path) {
@@ -56,12 +111,21 @@ static SSL_CTX* create_ssl_ctx(bool is_server, const char* cert, const char* key
 #ifdef SSL_OP_NO_RENEGOTIATION
   SSL_CTX_set_options(ctx, SSL_OP_NO_RENEGOTIATION);
 #endif
+  /* Let the server's own preference order decide the negotiated cipher rather
+   * than the client's, so a client cannot steer both peers into a weaker (but
+   * still offered) suite. */
+  SSL_CTX_set_options(ctx, SSL_OP_CIPHER_SERVER_PREFERENCE);
 
   if (SSL_CTX_set_min_proto_version(ctx, TLS1_2_VERSION) != 1) {
     SSL_CTX_free(ctx);
     return NULL;
   }
-  if (SSL_CTX_set_cipher_list(ctx, "HIGH:!aNULL:!eNULL:!MD5:!RC4:!3DES") != 1) {
+  /* TLS 1.2 and below: an AEAD-only suite list.  "HIGH" still includes CBC
+   * suites (Lucky13/POODLE-adjacent MAC-then-encrypt constructions), so restrict
+   * the list to ECDHE key agreement with an AEAD record cipher (AES-GCM or
+   * ChaCha20-Poly1305).  A NULL/weak/3DES cipher is never selectable. */
+  if (SSL_CTX_set_cipher_list(ctx, "ECDHE+AESGCM:ECDHE+CHACHA20:!aNULL:!eNULL:!MD5:!RC4:!3DES") !=
+      1) {
     SSL_CTX_free(ctx);
     return NULL;
   }
@@ -79,13 +143,6 @@ static SSL_CTX* create_ssl_ctx(bool is_server, const char* cert, const char* key
 #endif
 
   if (cert && key) {
-    struct stat key_stat;
-    if (stat(key, &key_stat) != 0 || !S_ISREG(key_stat.st_mode) || key_stat.st_uid != geteuid() ||
-        (key_stat.st_mode & (S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH))) {
-      log_message(LOG_LEVEL_ERROR, "TLS private key must be owned by the current user and private");
-      SSL_CTX_free(ctx);
-      return NULL;
-    }
     if (SSL_CTX_use_certificate_file(ctx, cert, SSL_FILETYPE_PEM) <= 0) {
       char* escaped = output_escape(cert, false);
       log_message(LOG_LEVEL_ERROR, "Failed to load certificate: %s",
@@ -95,12 +152,7 @@ static SSL_CTX* create_ssl_ctx(bool is_server, const char* cert, const char* key
       SSL_CTX_free(ctx);
       return NULL;
     }
-    if (SSL_CTX_use_PrivateKey_file(ctx, key, SSL_FILETYPE_PEM) <= 0) {
-      char* escaped = output_escape(key, false);
-      log_message(LOG_LEVEL_ERROR, "Failed to load private key: %s",
-                  escaped ? escaped : "<allocation failed>");
-      free(escaped);
-      log_ssl_errors();
+    if (!load_private_key_secure(ctx, key)) {
       SSL_CTX_free(ctx);
       return NULL;
     }
@@ -144,7 +196,16 @@ static SSL* wrap_fd_with_ssl(int fd, SSL_CTX* ctx, bool is_server, const char* h
   // Enable hostname verification for client connections when a hostname is provided.
   // Must be done before SSL_connect to take effect during the handshake.
   if (!is_server && hostname) {
-    if (SSL_set1_host(ssl, hostname) != 1) {
+    /* An IP-literal host must be verified against the certificate's IP SAN
+     * (X509_check_ip_asc), not as a DNS name: SSL_set1_host would look for a
+     * DNS SAN that a legitimate IP-SAN certificate never carries. */
+    struct in_addr ipv4;
+    struct in6_addr ipv6;
+    bool is_ip_literal =
+        inet_pton(AF_INET, hostname, &ipv4) == 1 || inet_pton(AF_INET6, hostname, &ipv6) == 1;
+    int set_ok = is_ip_literal ? X509_VERIFY_PARAM_set1_ip_asc(SSL_get0_param(ssl), hostname)
+                               : SSL_set1_host(ssl, hostname);
+    if (set_ok != 1) {
       SSL_free(ssl);
       return NULL;
     }

src/shared/utils.c

@@ -378,59 +378,155 @@ char* output_escape(const char* string, bool eight_bit_output) {
   return escaped;
 }
 
+ssize_t utils_getdelim_bounded(FILE* stream, char** line, size_t* cap, int delim, size_t max_len) {
+  if (!stream || !line || !cap || max_len == 0) {
+    errno = EINVAL;
+    return -1;
+  }
+  size_t limit = max_len + 1; /* content bytes plus the terminating NUL */
+  if (*line == NULL || *cap < 2) {
+    size_t initial = limit < 256 ? limit : 256;
+    char* buf = malloc(initial);
+    if (!buf)
+      return -1;
+    free(*line);
+    *line = buf;
+    *cap = initial;
+  }
+  size_t len = 0;
+  int c;
+  while ((c = getc_unlocked(stream)) != EOF) {
+    if (len >= max_len) {
+      errno = EFBIG;
+      return -1;
+    }
+    if (len + 2 > *cap) {
+      size_t new_cap = *cap * 2;
+      if (new_cap < len + 2)
+        new_cap = len + 2;
+      if (new_cap > limit)
+        new_cap = limit;
+      char* grown = realloc(*line, new_cap);
+      if (!grown)
+        return -1;
+      *line = grown;
+      *cap = new_cap;
+    }
+    (*line)[len++] = (char)c;
+    if (c == delim)
+      break;
+  }
+  if (c == EOF && len == 0)
+    return 0;
+  (*line)[len] = '\0';
+  return (ssize_t)len;
+}
+
 /* Match a glob pattern against a string. Supported wildcards:
  *   ?      matches any single character except '/'.
  *   *      matches any sequence of characters within one path component (no '/').
  *   **     matches any sequence of characters, including '/' (cross-directory).
  *   slash-star-star-slash is treated as a cross-directory wildcard when it appears between
  * literals.
- */
+ *
+ * The matcher is an iterative O(pattern * string) dynamic program rather than the
+ * original backtracking recursion: overlapping `*`/`**` wildcards made a pattern
+ * like `*a*a*a*...*b` run in exponential time against a long run of `a`, a CPU
+ * denial-of-service vector reachable from a hostile --exclude/--include pattern
+ * or `.rsync-filter`.  The DP reasons over (pattern position, string position)
+ * so every state is visited once; the transitions below mirror the original
+ * recursion exactly. */
 bool glob_match(const char* pattern, const char* str) {
-  while (*pattern) {
-    if (*pattern == '*') {
-      if (*(pattern + 1) == '*') {
-        /* globstar: match across directories */
-        pattern += 2;
-        if (*pattern == '\0')
-          return true;
-        if (*pattern == '/')
-          pattern++;
-        while (*str) {
-          if (glob_match(pattern, str))
-            return true;
-          str++;
+  if (!pattern || !str)
+    return false;
+  size_t pattern_len = strlen(pattern);
+  size_t str_len = strlen(str);
+  if (pattern_len == 0)
+    return str_len == 0;
+  /* Defensive work cap: the DP is bounded by pattern*string states, but a
+   * 64 KiB pattern against a 64 KiB path would still cost billions of steps.
+   * Treat the pattern as non-matching above the cap instead of burning CPU. */
+  if (str_len > (SIZE_MAX / (pattern_len + 1)) - 1)
+    return false;
+  if ((pattern_len + 1) * (str_len + 1) > 64u * 1024u * 1024u)
+    return false;
+
+  size_t row_bytes = str_len + 1;
+  /* Rows for pattern positions i, i+1, i+2 and i+3 are live at once (the
+   * globstar transition can skip up to three pattern bytes).  Four rotating
+   * rows keep memory at O(string length); a stack buffer avoids an allocation
+   * for the common short-leaf case. */
+  enum { STACK_ROW = 257 };
+  uint8_t stack_rows[4 * STACK_ROW];
+  uint8_t* rows = stack_rows;
+  if (row_bytes > STACK_ROW) {
+    rows = malloc(4 * row_bytes);
+    if (!rows)
+      return false;
+  }
+
+#define GLOB_ROW(i) (rows + ((pattern_len - (i)) & 3) * row_bytes)
+
+  /* Base row: pattern position `pattern_len` matches only the string's end. */
+  for (size_t j = 0; j <= str_len; j++)
+    GLOB_ROW(pattern_len)[j] = (j == str_len) ? 1 : 0;
+
+  for (size_t i = pattern_len; i-- > 0;) {
+    const char pc = pattern[i];
+    uint8_t* cur = GLOB_ROW(i);
+    const uint8_t* next = GLOB_ROW(i + 1);
+    if (pc == '*') {
+      if (i + 1 < pattern_len && pattern[i + 1] == '*') {
+        /* Globstar: skip `**` and an optional following '/', then consume any
+         * (possibly empty) run of characters -- including '/'. */
+        size_t rest = i + 2;
+        if (rest < pattern_len && pattern[rest] == '/')
+          rest++;
+        const uint8_t* rest_row = GLOB_ROW(rest);
+        for (size_t j = str_len + 1; j-- > 0;) {
+          bool v = rest_row[j] != 0;
+          if (!v && j < str_len)
+            v = cur[j + 1] != 0;
+          cur[j] = v ? 1 : 0;
+        }
+      } else {
+        /* Single `*`: zero characters, or one non-'/' character. */
+        for (size_t j = str_len + 1; j-- > 0;) {
+          bool v = next[j] != 0;
+          if (!v && j < str_len && str[j] != '/')
+            v = cur[j + 1] != 0;
+          cur[j] = v ? 1 : 0;
         }
-        return glob_match(pattern, str);
       }
-      /* single *: match within one path component */
-      pattern++;
-      while (*str && *str != '/') {
-        if (glob_match(pattern, str))
-          return true;
-        str++;
+    } else if (pc == '?') {
+      for (size_t j = str_len + 1; j-- > 0;) {
+        bool v = j < str_len && str[j] != '/' && next[j + 1] != 0;
+        cur[j] = v ? 1 : 0;
       }
-      return glob_match(pattern, str);
-    } else if (*pattern == '?') {
-      if (!*str || *str == '/')
-        return false;
-      pattern++;
-      str++;
     } else {
-      if (*pattern != *str) {
-        /* allow literal / ** / rest to match any number of directories */
-        if (*pattern == '/' && *(pattern + 1) == '*' && *(pattern + 2) == '*') {
-          const char* rest = pattern + 3;
-          if (*rest == '/')
+      /* Literal: consume an equal character, or -- for a '/' immediately before
+       * a globstar -- let the '/' match zero directories and continue at `**`. */
+      for (size_t j = str_len + 1; j-- > 0;) {
+        bool v = false;
+        if (j < str_len && str[j] == pc) {
+          v = next[j + 1] != 0;
+        } else if (pc == '/' && i + 2 < pattern_len && pattern[i + 1] == '*' &&
+                   pattern[i + 2] == '*') {
+          size_t rest = i + 3;
+          if (rest < pattern_len && pattern[rest] == '/')
             rest++;
-          return glob_match(rest, str);
+          v = GLOB_ROW(rest)[j] != 0;
         }
-        return false;
+        cur[j] = v ? 1 : 0;
       }
-      pattern++;
-      str++;
     }
   }
-  return *str == '\0';
+
+  bool matched = GLOB_ROW(0)[0] != 0;
+#undef GLOB_ROW
+  if (rows != stack_rows)
+    free(rows);
+  return matched;
 }
 
 bool format_human_bytes(unsigned long long bytes, char* buffer, size_t buffer_size) {

src/shared/utils.h

@@ -4,7 +4,9 @@
 #include "array_list.h"
 #include <stddef.h>
 #include <stdbool.h>
+#include <stdio.h>
 #include <sys/socket.h>
+#include <sys/types.h>
 
 /* Small open-addressing string hash set used to turn quadratic membership
  * scans into O(path length) exact-match lookups (the --delete keep-set and the
@@ -79,6 +81,17 @@ bool path_index_has_descendant(const PathIndex* index, const char* path);
 
 char* str_dup(const char* string);
 char* output_escape(const char* string, bool eight_bit_output);
+/* Upper bound on one line/token read from a local list file (--files-from,
+ * --exclude-from/--include-from, .rsync-filter).  Mirrors MAX_STRING_SIZE and
+ * stops a hostile multi-gigabyte line from forcing unbounded allocation. */
+#define UTILS_MAX_LINE_LEN (64 * 1024)
+/* Read one `delim`-terminated record from `stream` into *line (grown as needed
+ * and NUL-terminated), refusing to consume/allocate more than `max_len` bytes
+ * of content.  Returns the number of bytes stored (delimiter included, matching
+ * getdelim), 0 at end of file, or -1 on error (errno is EFBIG when the record
+ * exceeds `max_len`, ENOMEM on allocation failure).  *line and *cap are updated
+ * as the buffer grows and the caller owns *line. */
+ssize_t utils_getdelim_bounded(FILE* stream, char** line, size_t* cap, int delim, size_t max_len);
 char* path_cat(const char* path1, const char* path2);
 bool glob_match(const char* pattern, const char* str);
 /* Result of a bounded extra-file deletion run. */
@@ -148,6 +161,12 @@ const char* utils_get_authorized_root_path(void);
  * callers guarantee this); this is containment by string, not by resolved
  * symlinks.  Shared by the utils and file secure-walk root confinement. */
 bool path_is_within_root(const char* root, const char* path);
+/* True when `path` contains a ".." component.  This is a purely lexical
+ * dot-dot check: an absolute path is NOT rejected here, because default
+ * (non-relative) transfers legitimately put the sender's absolute source path
+ * on the wire and the receiver re-roots it under the destination with
+ * path_cat().  Callers that accept a strictly relative path (e.g. batch paths)
+ * must reject a leading '/' themselves (see utils_valid_batch_path). */
 bool has_path_traversal(const char* path);
 bool utils_valid_batch_path(const char* path);
 bool format_human_bytes(unsigned long long bytes, char* buffer, size_t buffer_size);

src/shared/xattr.c

@@ -73,13 +73,17 @@ bool xattr_list_append(FileXattrList* list, const char* name, const void* value,
 
 /* A Linux xattr name is "namespace.name" with an optional leading "trusted.",
  * "system.", "security.", "user.", or "trusted." prefix.  We only ever touch
- * the unprivileged "user.*" namespace and the two POSIX ACL xattrs carried in
- * the "system." namespace.  Everything else -- especially "security.*" (ACLs,
- * capabilities, SELinux labels) and "trusted.*" -- is refused so a client can
- * never compel the receiver to apply a privileged attribute it would not
- * otherwise be able to set (and which would be a local privilege escalation if
- * it could). */
-bool xattr_name_appliable(const char* name) {
+ * the unprivileged "user.*" namespace and, only when --acls/-A was negotiated,
+ * the two POSIX ACL xattrs carried in the "system." namespace.  Everything else
+ * -- especially "security.*" (ACLs, capabilities, SELinux labels) and
+ * "trusted.*" -- is refused so a client can never compel the receiver to apply a
+ * privileged attribute it would not otherwise be able to set (and which would be
+ * a local privilege escalation if it could).
+ *
+ * The ACL gate is deliberate: --xattrs/-X alone derives use_xattrs but must NOT
+ * authorize the ACL names, otherwise a -X client could plant an ACL the
+ * receiver never opted into (B4). */
+bool xattr_name_appliable(const char* name, bool preserve_acls) {
   if (!name || name[0] == '\0')
     return false;
   size_t len = strlen(name);
@@ -95,15 +99,24 @@ bool xattr_name_appliable(const char* name) {
   if (strncmp(name, "user.", 5) == 0)
     return name[5] != '\0';
   if (strcmp(name, "system.posix_acl_access") == 0)
-    return true;
+    return preserve_acls;
   if (strcmp(name, "system.posix_acl_default") == 0)
-    return true;
+    return preserve_acls;
   return false;
 }
 
+/* The two POSIX ACL xattr names: the only names whose applicablity is
+ * conditional (they require --acls).  Used by the receiver to distinguish "not
+ * negotiated" (drop the entry, keep user.* working for -X) from a genuinely
+ * disallowed namespace (hard reject). */
+static bool xattr_name_is_posix_acl(const char* name) {
+  return name != NULL && (strcmp(name, "system.posix_acl_access") == 0 ||
+                          strcmp(name, "system.posix_acl_default") == 0);
+}
+
 /* ---- SENDER: capture ---- */
 
-FileXattrList* xattr_capture_path(const char* path) {
+FileXattrList* xattr_capture_path(const char* path, bool preserve_acls) {
   if (!path)
     return NULL;
   ssize_t list_size = listxattr(path, NULL, 0);
@@ -130,7 +143,10 @@ FileXattrList* xattr_capture_path(const char* path) {
     if (name_len == 0)
       break; /* trailing double NUL not expected; stop */
     offset += (ssize_t)name_len + 1;
-    if (!xattr_name_appliable(name))
+    /* Capture is sender-side: the scanner has already gated on -X/-A, so the
+       per-name whitelist here allows the ACL names only when --acls was
+       negotiated.  Without it a plain -X capture never carries an ACL. */
+    if (!xattr_name_appliable(name, preserve_acls))
       continue;
     ssize_t value_size = getxattr(path, name, NULL, 0);
     if (value_size < 0)
@@ -190,7 +206,7 @@ bool xattr_send(int fd, const FileXattrList* list) {
   return true;
 }
 
-FileXattrList* xattr_receive(int fd, int* ok) {
+FileXattrList* xattr_receive(int fd, int* ok, bool preserve_acls) {
   if (ok)
     *ok = 0;
   int count;
@@ -232,11 +248,19 @@ FileXattrList* xattr_receive(int fd, int* ok) {
       xattr_list_free(list);
       return NULL;
     }
-    if (!xattr_name_appliable(name)) {
-      log_message(LOG_LEVEL_ERROR, "rejected xattr block: disallowed namespace for '%s'", name);
-      free(name);
-      xattr_list_free(list);
-      return NULL;
+    bool skip = false;
+    if (!xattr_name_appliable(name, preserve_acls)) {
+      if (!preserve_acls && xattr_name_is_posix_acl(name)) {
+        /* -X without -A: the sender may still carry ACLs, but the receiver must
+           never apply an ACL it was not asked to preserve.  Consume and drop the
+           entry (keeping -X compatibility) rather than failing the transfer. */
+        skip = true;
+      } else {
+        log_message(LOG_LEVEL_ERROR, "rejected xattr block: disallowed namespace for '%s'", name);
+        free(name);
+        xattr_list_free(list);
+        return NULL;
+      }
     }
     int32_t value_len32;
     if (!receive_n_data(fd, &value_len32, sizeof(value_len32))) {
@@ -273,6 +297,12 @@ FileXattrList* xattr_receive(int fd, int* ok) {
         return NULL;
       }
     }
+    if (skip) {
+      free(value);
+      free(name);
+      budget += (size_t)name_len32 + (size_t)value_len32;
+      continue;
+    }
     if (!xattr_list_append(list, name, value, (size_t)value_len32)) {
       free(value);
       free(name);

src/shared/xattr.h

@@ -59,20 +59,28 @@ void xattr_list_free(FileXattrList* list);
 bool xattr_list_append(FileXattrList* list, const char* name, const void* value, size_t value_len);
 
 /* True when `name` is a well-formed xattr name AND belongs to a namespace this
- * build is authorized to apply (user.* or the two POSIX ACL xattrs).  Used for
- * both capture and receiver-side validation. */
-bool xattr_name_appliable(const char* name);
+ * build is authorized to apply.  `user.*` is always accepted for -X; the two
+ * POSIX ACL xattrs are accepted only when `preserve_acls` (--acls/-A) is set, so
+ * a plain -X run can never carry or apply an ACL the receiver did not ask for.
+ * Used for both capture and receiver-side validation. */
+bool xattr_name_appliable(const char* name, bool preserve_acls);
 
-/* Sender: read the whitelisted xattrs of `path` into a new list.  Returns NULL
- * when the path has no appliable xattrs (or the filesystem has no xattr
- * support); an empty-but-valid list is never returned distinct from NULL. */
-FileXattrList* xattr_capture_path(const char* path);
+/* Sender: read the whitelisted xattrs of `path` into a new list.  The POSIX ACL
+ * names are captured only when `preserve_acls` (--acls/-A) is set, so a plain
+ * -X run never carries an ACL it was not asked to preserve; `user.*` is
+ * unaffected.  Returns NULL when the path has no appliable xattrs (or the
+ * filesystem has no xattr support); an empty-but-valid list is never returned
+ * distinct from NULL. */
+FileXattrList* xattr_capture_path(const char* path, bool preserve_acls);
 
 /* Wire: bounded serialization.  xattr_send returns false on write failure; an
  * empty/NULL list transmits a zero-count block.  xattr_receive returns NULL and
- * sets *ok = 0 on any malformed / oversized / non-whitelisted entry. */
+ * sets *ok = 0 on any malformed / oversized / non-whitelisted entry.  When
+ * `preserve_acls` is false, any POSIX ACL entries are consumed and DROPPED (so
+ * a -X transfer still succeeds and never applies an ACL it did not negotiate);
+ * a genuinely disallowed namespace is still rejected. */
 bool xattr_send(int fd, const FileXattrList* list);
-FileXattrList* xattr_receive(int fd, int* ok);
+FileXattrList* xattr_receive(int fd, int* ok, bool preserve_acls);
 
 /* Receiver: apply every entry fd-relative (fsetxattr) to the just-written file
  * descriptor.  A per-attribute failure (e.g. ACL set refused for non-root on a

tests/conftest.py

@@ -16,7 +16,13 @@ def shared_server():
     Under pytest-xdist this session fixture is instantiated once per worker
     process, so each worker gets its own server on an ephemeral port."""
     server = ServerManager()
-    server.start()
+    # --allow-super keeps the historical permissive super mode for a root
+    # receiver: the integration suite's root-only ownership/device/copy-as tests
+    # exercise that opted-in configuration.  The secure default (a root
+    # standalone server without --allow-super forces SUPER_MODE_OFF) is covered
+    # explicitly by TestStandaloneSuperDefault in test_features.py.  Non-root
+    # runs are unaffected by the flag.
+    server.start(extra_args=["--allow-super"])
     yield server
     server.stop()
 

tests/fuzz/fuzz_xattr_block.c

@@ -75,7 +75,7 @@ static void write_best_effort(int fd, const void* data, size_t size) {
 }
 
 static void receive_stream(const unsigned char* prefix, size_t prefix_len, const uint8_t* data,
-                           size_t size) {
+                           size_t size, bool preserve_acls) {
   int sv[2];
   if (socketpair(AF_UNIX, SOCK_STREAM, 0, sv) != 0)
     return;
@@ -91,7 +91,7 @@ static void receive_stream(const unsigned char* prefix, size_t prefix_len, const
   shutdown(sv[0], SHUT_WR);
 
   int ok = 0;
-  FileXattrList* list = xattr_receive(sv[1], &ok);
+  FileXattrList* list = xattr_receive(sv[1], &ok, preserve_acls);
   xattr_list_free(list);
 
   close(sv[0]);
@@ -102,14 +102,18 @@ int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
   if (!g_block_ready)
     build_canonical_block();
 
-  /* Raw bytes as the whole block. */
-  receive_stream(NULL, 0, data, size);
+  /* Raw bytes as the whole block.  Exercise both the -X-only (no ACLs) and the
+   * -A (ACL names accepted) receiver gates. */
+  for (int acls = 0; acls < 2; acls++) {
+    bool preserve_acls = acls != 0;
+    receive_stream(NULL, 0, data, size, preserve_acls);
 
-  /* Valid framing so the fuzzer mutates the entry list, the first value and
-   * the second entry respectively instead of stopping at the count. */
-  receive_stream(g_block, g_off_after_entry0, data, size);
-  receive_stream(g_block, g_off_value0, data, size);
-  receive_stream(g_block, g_off_after_count, data, size);
+    /* Valid framing so the fuzzer mutates the entry list, the first value and
+     * the second entry respectively instead of stopping at the count. */
+    receive_stream(g_block, g_off_after_entry0, data, size, preserve_acls);
+    receive_stream(g_block, g_off_value0, data, size, preserve_acls);
+    receive_stream(g_block, g_off_after_count, data, size, preserve_acls);
+  }
 
   return 0;
 }

tests/integration/test_batch.py

@@ -117,4 +117,16 @@ def test_batch_modes_conflict():
         cmd = _run(["--source-dir", SOURCE_DIR, "--dest-dir", DEST1] + flags)
         result = subprocess.run(cmd, capture_output=True, text=True, timeout=180)
         assert result.returncode != 0, \
-            f"expected conflict failure for {flags}: {result.stderr}"
+            f"expected conflict failure for {flags}: {result.stderr}"
+
+
+@pytest.mark.ci
+def test_dry_run_rejects_write_batch():
+    """--dry-run must not emit a batch file (it must not mutate anything)."""
+    if os.path.exists(BATCH_FILE):
+        os.unlink(BATCH_FILE)
+    cmd = _run(["--source-dir", SOURCE_DIR, "--dest-dir", DEST1,
+                "--dry-run", "--write-batch", BATCH_FILE])
+    result = subprocess.run(cmd, capture_output=True, text=True, timeout=180)
+    assert result.returncode != 0, result.stderr
+    assert not os.path.exists(BATCH_FILE), "dry-run must not create a batch file"

tests/integration/test_features.py

@@ -444,9 +444,10 @@ class TestRemoteDryRun:
         self._seed(source)
         clean_dir(dest)
 
-        # Populate the destination with a real transfer, then make exactly one
-        # file differ (content+size) and add a brand-new file.
-        result, _ = run_client(source, dest, port=shared_server.port)
+        # Populate the destination with a real transfer that preserves mtimes
+        # (--preserve), then make exactly one file differ (content+size) and add
+        # a brand-new file.
+        result, _ = run_client(source, dest, flags=["--preserve"], port=shared_server.port)
         assert result.returncode == 0, f"seed transfer failed: {result.stderr[:200]}"
         received = get_dest_received_dir(dest, source)
 
@@ -456,9 +457,11 @@ class TestRemoteDryRun:
             f.write(b"newly added\n")
 
         before = _snapshot_tree(received)
-        # --checksum makes the up-to-date decision content-based (the seed
-        # transfer did not preserve mtimes), so keep.txt/deep.txt report skip.
-        result, _ = run_client(source, dest, flags=["--dry-run", "--checksum"],
+        # --checksum must NOT read destination contents in a dry-run (B3), so
+        # the up-to-date decision is metadata-only.  The --preserve seed made
+        # keep.txt and deep.txt size+mtime-identical; the dry-run must also
+        # transmit metadata (--preserve) for that metadata to be comparable.
+        result, _ = run_client(source, dest, flags=["--dry-run", "--checksum", "--preserve"],
                                port=shared_server.port)
         assert result.returncode == 0, f"remote dry-run failed: {result.stderr[:300]}"
         assert "Dry run:" in result.stdout, result.stdout[:200]
@@ -470,6 +473,34 @@ class TestRemoteDryRun:
         assert "deep.txt" not in result.stdout, result.stdout
         assert _snapshot_tree(received) == before, "remote dry-run mutated the destination"
 
+    @pytest.mark.ci
+    def test_remote_dry_run_checksum_does_not_read_destination(self, shared_server):
+        """B3: --dry-run --checksum against a read-only module must not read the
+        destination file's content (a 1-bit hash oracle).  A same-size/same-content
+        file whose mtime differs is therefore reported as would-transfer because
+        the metadata-only decision is inconclusive, instead of being hashed and
+        silently skipped."""
+        source = os.path.join(TEST_DATA_DIR, "remote_dry_oracle_src")
+        dest = os.path.join(TEST_DATA_DIR, "remote_dry_oracle_dst")
+        self._seed(source)
+        clean_dir(dest)
+        result, _ = run_client(source, dest, flags=["--preserve"], port=shared_server.port)
+        assert result.returncode == 0, result.stderr[:200]
+        received = get_dest_received_dir(dest, source)
+
+        target = os.path.join(received, "keep.txt")
+        # Identical size and content, but a deliberately different mtime.
+        os.utime(target, (1000000000, 1000000000))
+        before = _snapshot_tree(received)
+
+        result, _ = run_client(source, dest, flags=["--dry-run", "--checksum", "--preserve"],
+                               port=shared_server.port)
+        assert result.returncode == 0, result.stderr[:300]
+        assert "keep.txt" in result.stdout, (
+            f"dry-run --checksum must not read the destination to prove equality: {result.stdout}"
+        )
+        assert _snapshot_tree(received) == before, "dry-run mutated the destination"
+
     @pytest.mark.ci
     def test_remote_dry_run_into_empty_dest_creates_nothing(self, shared_server):
         source = os.path.join(TEST_DATA_DIR, "remote_dry_empty_src")
@@ -1561,8 +1592,33 @@ class TestDelete:
         assert not missing, f"Missing: {missing}"
         assert not mismatches, f"Mismatch: {mismatches}"
 
-
-class TestProgress:
+    @pytest.mark.ci
+    def test_force_cannot_replace_directory_without_allow_delete(self):
+        """C2: --force is deletion authority (an incoming file may recursively
+        remove a non-empty destination directory tree).  A server started without
+        --allow-delete must clear it, so the operator's delete policy cannot be
+        bypassed with --force."""
+        source = os.path.join(TEST_DATA_DIR, "force_src")
+        dest = os.path.join(TEST_DATA_DIR, "force_dst")
+        clean_dir(source)
+        clean_dir(dest)
+        with open(os.path.join(source, "blocker"), "wb") as f:
+            f.write(b"incoming file\n")
+        received = get_dest_received_dir(dest, source)
+        blocker = os.path.join(received, "blocker")
+        os.makedirs(blocker)
+        nested = os.path.join(blocker, "nested.txt")
+        with open(nested, "w") as f:
+            f.write("survivor")
+        # Deliberately NO --allow-delete.
+        server = ServerManager()
+        server.start()
+        try:
+            run_client(source, dest, flags=["--force"], port=server.port)
+        finally:
+            server.stop()
+        assert os.path.isdir(blocker), "unauthorized --force removed a destination directory"
+        assert os.path.exists(nested), "unauthorized --force removed a nested file"
     def test_progress_output(self, shared_server):
         clean_dir(DEST_DIR)
         result, dur = run_client(
@@ -3782,6 +3838,27 @@ class TestBasisDestDirs:
         assert _read_file(os.path.join(received, self.ADDED)) == \
             self._source_tree("c")[self.ADDED], "added file not transferred"
 
+    @pytest.mark.ci
+    def test_dry_run_compare_dest_does_not_read_basis(self, shared_server):
+        # A dry-run --compare-dest must never read/hash the basis file: doing so
+        # is a 1-bit content oracle against the client-supplied digest.  Even a
+        # byte-identical basis with a matching size+mtime is therefore reported
+        # as would-transfer, and nothing is created.
+        source = self._make_source("basis_dry_src", {self.UNCHANGED: b"stable content v1\n"})
+        dest = os.path.join(TEST_DATA_DIR, "basis_dry_dst")
+        clean_dir(dest)
+        self._seed_basis(dest, source, "drybasis", {self.UNCHANGED: b"stable content v1\n"})
+        before = _snapshot_tree(dest)
+        result, _ = run_client(source, dest,
+                               flags=["--compare-dest=drybasis", "--dry-run"],
+                               port=shared_server.port)
+        assert result.returncode == 0, \
+            f"dry-run compare-dest failed: {result.stderr[:300]}"
+        assert self.UNCHANGED in result.stdout, (
+            "dry-run compare-dest silently skipped: receiver read the basis content"
+        )
+        assert _snapshot_tree(dest) == before, "dry-run compare-dest mutated the destination"
+
     def test_compare_dest_content_mismatch_forces_transfer(self, shared_server):
         # The basis holds a file with a DIFFERENT body: even though it shares
         # the mtime pin, the xxHash check fails and the data must be sent.
@@ -4578,6 +4655,61 @@ class TestSuperPrivilege:
             f"--no-super must suppress fake-super's owner replay: uid={st.st_uid} gid={st.st_gid}"
 
 
+class TestStandaloneSuperDefault:
+    """C3: a privileged (root) STANDALONE server without --allow-super forces
+    SUPER_MODE_OFF, so a client cannot make it create device nodes, write raw
+    devices, apply ownership, or use --copy-as.  The shared_server fixture opts in
+    with --allow-super to keep the historical behavior available to the existing
+    root-only tests; these tests start their own un-opted server."""
+
+    @pytest.mark.ci
+    def test_copy_as_refused_without_allow_super(self):
+        """--copy-as is a client-chosen-ownership request and must be refused by
+        a standalone server that did not opt in with --allow-super (on a non-root
+        receiver it is refused for lack of privilege either way)."""
+        source = os.path.join(TEST_DATA_DIR, "super_default_src")
+        dest = os.path.join(TEST_DATA_DIR, "super_default_dst")
+        clean_dir(source)
+        clean_dir(dest)
+        with open(os.path.join(source, "f.txt"), "wb") as f:
+            f.write(b"no copy-as\n")
+        server = ServerManager()
+        server.start()  # deliberately no --allow-super
+        try:
+            result, _ = run_client(source, dest,
+                                   flags=["--preserve", "--copy-as=@65534:@65534"],
+                                   port=server.port)
+        finally:
+            server.stop()
+        assert result.returncode != 0, (
+            "standalone server accepted --copy-as without --allow-super"
+        )
+
+    @pytest.mark.skipif(os.geteuid() != 0, reason="root can create the source device node")
+    def test_devices_skipped_without_allow_super(self):
+        """Root standalone server without --allow-super must skip device-node
+        creation even for a client --devices request (the run still succeeds and
+        the regular file transfers)."""
+        source = os.path.join(TEST_DATA_DIR, "super_default_dev_src")
+        dest = os.path.join(TEST_DATA_DIR, "super_default_dev_dst")
+        clean_dir(source)
+        clean_dir(dest)
+        with open(os.path.join(source, "plain.txt"), "wb") as f:
+            f.write(b"regular\n")
+        os.mknod(os.path.join(source, "null"), stat.S_IFCHR | 0o666, os.makedev(1, 3))
+        server = ServerManager()
+        server.start()  # deliberately no --allow-super
+        try:
+            result, _ = run_client(source, dest, flags=["--devices"], port=server.port)
+        finally:
+            server.stop()
+        assert result.returncode == 0, f"exit {result.returncode}: {(result.stderr or '')[:200]}"
+        received = get_dest_received_dir(dest, source)
+        assert not os.path.lexists(os.path.join(received, "null")), (
+            "root standalone server created a device node without --allow-super"
+        )
+
+
 class TestHardLinks:
     """-H/--hard-links: source files sharing an inode are re-created as hard
     links to one another on the destination (dedup preserved, first copy

tests/test_chunk.c

@@ -1,8 +1,10 @@
 
 #include "chunk.h"
+#include "protocol.h"
 #include "test_utils.h"
 #include "utils.h"
 #include <string.h>
+#include <sys/socket.h>
 #include <sys/stat.h>
 #include <unistd.h>
 
@@ -279,6 +281,51 @@ static void test_chunk_special_rdev_out_of_range_rejected() {
   chunk_destroy(chunk);
 }
 
+/* B6: chunk_deserialize() charges each retained per-file copy to the owning
+ * session's connection budget (MAX_CONNECTION_MEMORY) so queued chunk payloads
+ * are not held outside the per-connection ceiling; destroying the chunk returns
+ * the charge through the Data.owner path. */
+static void test_chunk_deserialize_charges_session_budget() {
+  const char* path = "temp_chunk_charge.txt";
+  const char* content = "charge me to the connection budget";
+  unlink(path);
+  file_write_to_disk(path, content, strlen(content), false, false);
+  struct stat st;
+  EXPECT_EQ_INT(stat(path, &st), 0);
+
+  int p[2];
+  EXPECT_EQ_INT(socketpair(AF_UNIX, SOCK_STREAM, 0, p), 0);
+  ProtocolSession session;
+  protocol_session_init(&session, p[0], p[1]);
+  protocol_session_set_max_alloc(&session, 4ULL * 1024 * 1024);
+
+  File* f = file_create(path);
+  EXPECT_NOT_NULL(f);
+  f->data->size = (unsigned long long)st.st_size;
+  EXPECT_TRUE(file_load_data(f));
+  File* files[1] = {f};
+  Chunk* chunk = chunk_create(files, 1);
+  EXPECT_NOT_NULL(chunk);
+  Data* serialized = chunk_serialize(chunk, false);
+  EXPECT_NOT_NULL(serialized);
+  /* Simulate a received buffer carrying its owning session. */
+  serialized->owner = &session;
+
+  Chunk* deserialized = chunk_deserialize(serialized, false);
+  EXPECT_NOT_NULL(deserialized);
+  unsigned long long charged = atomic_load(&session.total_allocated_bytes);
+  EXPECT_EQ_INT((int)charged, (int)strlen(content));
+  chunk_destroy(deserialized);
+  /* The copy's charge is released with the File/Data on destroy. */
+  EXPECT_EQ_INT((int)atomic_load(&session.total_allocated_bytes), 0);
+
+  data_destroy(serialized);
+  chunk_destroy(chunk);
+  close(p[0]);
+  close(p[1]);
+  unlink(path);
+}
+
 void test_chunk() {
   test_file_operations();
   test_chunk_operations();
@@ -286,4 +333,5 @@ void test_chunk() {
   test_chunk_symlink_roundtrip();
   test_chunk_special_rdev_roundtrip();
   test_chunk_special_rdev_out_of_range_rejected();
+  test_chunk_deserialize_charges_session_budget();
 }

tests/test_client_cli.c

@@ -3275,6 +3275,58 @@ static void test_parse_args_block_size() {
   config_delete(cfg);
 }
 
+/* An over-long --exclude-from/--include-from line is rejected at parse time
+ * rather than being read without a bound. */
+static void test_parse_args_pattern_file_oversized_rejected() {
+  const char* list_path = "cli_pattern_oversized.txt";
+  size_t len = UTILS_MAX_LINE_LEN + 4096;
+  char* big = malloc(len);
+  EXPECT_NOT_NULL(big);
+  memset(big, 'a', len);
+  write_file_bytes(list_path, big, len);
+  free(big);
+
+  Config* cfg = config_create();
+  int positional_args[2];
+  int positional_count = 0;
+  char* argv[] = {"fastsync", "--exclude-from", (char*)list_path, "/src", "/dst"};
+  EXPECT_EQ_INT(parse_args(cfg, 5, argv, positional_args, &positional_count), -1);
+  config_delete(cfg);
+  remove(list_path);
+}
+
+/* A leading '-'/'+' must be rejected for every unsigned numeric option so
+ * strtoull can never silently wrap (e.g. -1 -> ULLONG_MAX). */
+static void test_parse_args_unsigned_options_reject_sign() {
+  static const char* const opts[] = {"--chunk-size", "--bwlimit", "--delta-max"};
+  for (size_t i = 0; i < sizeof(opts) / sizeof(opts[0]); i++) {
+    Config* cfg = config_create();
+    int positional_args[2];
+    int positional_count = 0;
+    char* argv[] = {"fastsync", (char*)opts[i], "-1", "/src", "/dst"};
+    EXPECT_EQ_INT(parse_args(cfg, 5, argv, positional_args, &positional_count), -1);
+    config_delete(cfg);
+  }
+
+  /* An over-cap --chunk-size is rejected at parse time (max 64 MiB). */
+  Config* cfg = config_create();
+  int positional_args[2];
+  int positional_count = 0;
+  char* big_argv[] = {"fastsync", "--chunk-size", "67108865", "/src", "/dst"};
+  EXPECT_EQ_INT(parse_args(cfg, 5, big_argv, positional_args, &positional_count), -1);
+  config_delete(cfg);
+}
+
+/* --dry-run must not emit a batch file, so it is rejected alongside
+ * --read-batch/--only-write-batch. */
+static void test_validate_config_dry_run_rejects_write_batch() {
+  Config* cfg = valid_client_config();
+  cfg->dry_run = true;
+  cfg->write_batch = str_dup("batch.dat");
+  EXPECT_FALSE(validate_config(cfg));
+  config_delete(cfg);
+}
+
 void test_client_cli() {
   test_validate_config_required_paths();
   test_parse_args_numeric_ids();
@@ -3432,4 +3484,7 @@ void test_client_cli() {
   test_parse_args_remote_option_short_M();
   test_parse_args_no_motd();
   test_parse_args_password_file();
+  test_parse_args_pattern_file_oversized_rejected();
+  test_parse_args_unsigned_options_reject_sign();
+  test_validate_config_dry_run_rejects_write_batch();
 }

tests/test_compression.c

@@ -6,8 +6,10 @@
 #include "utils.h"
 #include <string.h>
 #include <sys/stat.h>
+#include <sys/wait.h>
 #include <threads.h>
 #include <unistd.h>
+#include <zstd.h>
 
 static void test_data_compress_decompress_roundtrip() {
   const char original[] = "Hello, World! This is test data for compression round-trip!";
@@ -138,6 +140,63 @@ static void test_chunk_compress_decompress_roundtrip() {
   unlink(path2);
 }
 
+/* Build a zstd frame whose header omits the content size (the content size
+ * flag is cleared), which ZSTD_getFrameContentSize reports as
+ * ZSTD_CONTENTSIZE_UNKNOWN. */
+static Data* make_unknown_size_frame(const void* src, size_t len) {
+  ZSTD_CCtx* cctx = ZSTD_createCCtx();
+  if (!cctx)
+    return NULL;
+  ZSTD_CCtx_setParameter(cctx, ZSTD_c_contentSizeFlag, 0);
+  size_t cap = ZSTD_compressBound(len);
+  Data* out = data_create_empty(cap);
+  if (!out) {
+    ZSTD_freeCCtx(cctx);
+    return NULL;
+  }
+  ZSTD_inBuffer in = {src, len, 0};
+  ZSTD_outBuffer ob = {out->data, cap, 0};
+  size_t ret;
+  do {
+    ret = ZSTD_compressStream2(cctx, &ob, &in, ZSTD_e_end);
+    if (ZSTD_isError(ret)) {
+      data_destroy(out);
+      ZSTD_freeCCtx(cctx);
+      return NULL;
+    }
+  } while (ret > 0);
+  out->size = ob.pos;
+  ZSTD_freeCCtx(cctx);
+  return out;
+}
+
+/* ZSTD_CONTENTSIZE_UNKNOWN is flagged by ZSTD_isError(), so a naive
+ * ZSTD_isError() check rejects every unknown-size frame.  Such a frame must
+ * instead reach the 3x estimate fallback and decompress correctly. */
+static void test_data_decompress_unknown_size_frame() {
+  const char original[] = "unknown-content-size frame: the decompressor must use the 3x estimate, "
+                          "not reject the frame as an error.";
+  size_t len = strlen(original);
+  char* buf = malloc(len);
+  EXPECT_NOT_NULL(buf);
+  memcpy(buf, original, len);
+
+  Data* frame = make_unknown_size_frame(buf, len);
+  free(buf);
+  EXPECT_NOT_NULL(frame);
+  /* Guard the premise of the test: the frame really has no stored size. */
+  EXPECT_EQ_INT((int)ZSTD_getFrameContentSize(frame->data, frame->size),
+                (int)ZSTD_CONTENTSIZE_UNKNOWN);
+
+  Data* decompressed = data_decompress(frame);
+  EXPECT_NOT_NULL(decompressed);
+  EXPECT_EQ_INT((int)decompressed->size, (int)len);
+  EXPECT_EQ_INT(memcmp(decompressed->data, original, len), 0);
+
+  data_destroy(decompressed);
+  data_destroy(frame);
+}
+
 typedef struct {
   int id;
   int iterations;
@@ -211,9 +270,52 @@ static void test_data_compress_reused_contexts_multithreaded() {
   compression_free_thread_contexts();
 }
 
+/* A truncated zstd frame used to make the decompressor spin forever: the
+ * stream call keeps returning a positive hint with all input consumed.  Run the
+ * decompression in a child with an alarm so a regression (infinite loop) is
+ * caught as a timeout failure instead of hanging the whole unit suite. */
+static void test_data_decompress_truncated_frame_fails() {
+  const char* original =
+      "The quick brown fox jumps over the lazy dog. The quick brown fox jumps over the lazy dog.";
+  size_t len = strlen(original);
+  char* buf = malloc(len);
+  EXPECT_NOT_NULL(buf);
+  memcpy(buf, original, len);
+  Data* input = data_create(buf, len);
+  EXPECT_NOT_NULL(input);
+
+  pid_t pid = fork();
+  EXPECT_TRUE(pid >= 0);
+  if (pid == 0) {
+    alarm(10); /* kills the child if the decompressor hangs */
+    Data* compressed = data_compress(input, 3);
+    if (compressed && compressed->size > 1) {
+      compressed->size -= 1; /* drop the final byte: frame is now incomplete */
+      Data* out = data_decompress(compressed);
+      bool failed_cleanly = (out == NULL);
+      data_destroy(out);
+      data_destroy(compressed);
+      /* The child inherited `input` across fork(); free it before _exit so the
+       * valgrind CI job (which instruments forked children too) sees no leak. */
+      data_destroy(input);
+      _exit(failed_cleanly ? 0 : 1);
+    }
+    data_destroy(compressed);
+    data_destroy(input);
+    _exit(2);
+  }
+  int status;
+  waitpid(pid, &status, 0);
+  EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+
+  data_destroy(input);
+}
+
 void test_compression() {
   test_data_compress_decompress_roundtrip();
   test_data_compress_decompress_large();
+  test_data_decompress_unknown_size_frame();
+  test_data_decompress_truncated_frame_fails();
   test_skip_compress_suffix_matching();
   test_data_compress_with_threads_roundtrip();
   test_data_compress_reused_contexts_multithreaded();

tests/test_config.c

@@ -87,6 +87,34 @@ static void test_config_ssh_dest_no_user() {
   config_delete(cfg);
 }
 
+/* C1: the user@host token is passed to ssh in option position, so a host or user
+ * beginning with '-' (e.g. "-oProxyCommand=...") must be rejected before any
+ * argv is built, and an empty host must be rejected too. */
+static void test_config_ssh_dest_rejects_option_injection() {
+  Config* cfg = make_config("1.0", "/src", "-oProxyCommand=id:/dst", true, false, false, false,
+                            false, 1, false, 0);
+  EXPECT_EQ_INT(config_parse_ssh_dest(cfg), -1);
+  EXPECT_EQ_INT(cfg->transport, TRANSPORT_TCP);
+  EXPECT_NULL(cfg->ssh_destination);
+  config_delete(cfg);
+
+  cfg =
+      make_config("1.0", "/src", "-user@host:/dst", true, false, false, false, false, 1, false, 0);
+  EXPECT_EQ_INT(config_parse_ssh_dest(cfg), -1);
+  config_delete(cfg);
+
+  cfg = make_config("1.0", "/src", "user@:/dst", true, false, false, false, false, 1, false, 0);
+  EXPECT_EQ_INT(config_parse_ssh_dest(cfg), -1);
+  config_delete(cfg);
+
+  /* config_parse_transport_dest propagates the rejection (and still returns 1
+   * for daemon syntax first). */
+  cfg = make_config("1.0", "/src", "-oProxyCommand=id:/dst", true, false, false, false, false, 1,
+                    false, 0);
+  EXPECT_EQ_INT(config_parse_transport_dest(cfg), -1);
+  config_delete(cfg);
+}
+
 static void test_config_daemon_dest_parse() {
   Config* cfg = make_config("1.0", "/src", "dahost::files/sub/dir", true, false, false, false,
                             false, 1, false, 0);
@@ -2789,6 +2817,7 @@ void test_config() {
   test_config_ssh_dest();
   test_config_ssh_dest_local_path();
   test_config_ssh_dest_no_user();
+  test_config_ssh_dest_rejects_option_injection();
   test_config_daemon_dest_parse();
   test_config_daemon_dest_no_path();
   test_config_daemon_dest_double_slash_normalized();

tests/test_credentials.c

@@ -473,6 +473,34 @@ static void test_credentials_store_rejects_legacy_hex() {
   free(path);
 }
 
+/* C9: the legacy-hex detector must check the length before indexing 64 bytes, so
+ * a short secret is never read out of bounds.  Such a line is rejected for the
+ * ordinary "expected verifier" reason, never as legacy. */
+static void test_credentials_store_rejects_short_secret() {
+  char contents[CREDENTIAL_MAX_LINE];
+  snprintf(contents, sizeof(contents), "alice:%s\n", "abc");
+  char* path = make_tmp_file(contents);
+  EXPECT_NOT_NULL(path);
+  char err[512];
+  const CredentialStore* store = credentials_load(path, NULL, err, sizeof(err));
+  EXPECT_NULL(store);
+  EXPECT_TRUE(strstr(err, "legacy unsalted") == NULL);
+  rm_temp(path);
+  free(path);
+
+  char short_hex[64];
+  memset(short_hex, 'a', 63);
+  short_hex[63] = '\0';
+  snprintf(contents, sizeof(contents), "alice:%s\n", short_hex);
+  path = make_tmp_file(contents);
+  EXPECT_NOT_NULL(path);
+  store = credentials_load(path, NULL, err, sizeof(err));
+  EXPECT_NULL(store);
+  EXPECT_TRUE(strstr(err, "legacy unsalted") == NULL);
+  rm_temp(path);
+  free(path);
+}
+
 static void test_credentials_store_duplicate_rejected() {
   char line[CREDENTIAL_MAX_LINE];
   EXPECT_TRUE(make_store_line("alice", KAT_PASSWORD, CREDENTIAL_MIN_ITERS, line, sizeof(line)));
@@ -1066,6 +1094,7 @@ void test_credentials(void) {
   test_credentials_store_parse_valid();
   test_credentials_store_parse_rejects_malformed();
   test_credentials_store_rejects_legacy_hex();
+  test_credentials_store_rejects_short_secret();
   test_credentials_store_duplicate_rejected();
   test_credentials_store_rejects_nonuniform_iters();
   test_credentials_store_parse_missing_file();

tests/test_daemon_conf.c

@@ -391,6 +391,22 @@ static void test_daemon_conf_auth_users_validated() {
   EXPECT_EQ_STR(ok_conf->modules[0].auth_users[0], "alice");
   EXPECT_EQ_STR(ok_conf->modules[0].auth_users[1], "bob");
   daemon_conf_free(ok_conf);
+
+  /* C4: an empty or separator-only `auth users` value is a parse error.  It
+   * would otherwise leave the module with a zero-length allow-list, silently
+   * disabling the authentication the operator asked for. */
+  const char* empty_auth[] = {
+      "[m]\npath = /x\nauth users = \n",
+      "[m]\npath = /x\nauth users = , ,\n",
+      "[m]\npath = /x\nauth users = \t\n",
+  };
+  for (size_t i = 0; i < sizeof(empty_auth) / sizeof(empty_auth[0]); i++) {
+    EXPECT_EQ_INT(write_conf(empty_auth[i], &path), 0);
+    const DaemonConf* rejected = daemon_conf_load(path, err, sizeof(err));
+    free(path);
+    EXPECT_NULL(rejected);
+    EXPECT_TRUE(strstr(err, "'auth users' must list at least one user") != NULL);
+  }
 }
 
 /* Wave 3 daemon hardening: configurable global/per-module connection caps,
@@ -475,12 +491,55 @@ static void test_daemon_conf_limits_and_hosts_parse() {
   EXPECT_EQ_INT(conf->modules[0].max_connections, 0);
   daemon_conf_free(conf);
 
-  /* An empty hosts list is not an error (no patterns are added). */
-  EXPECT_EQ_INT(write_conf("hosts allow = \n[m]\npath = /x\n", &path), 0);
+  /* C4: a present hosts key with an empty/separator-only value must not silently
+   * install a zero-length (allow-everyone) list. */
+  const char* empty_hosts[] = {
+      "hosts allow = \n[m]\npath = /x\n",
+      "hosts deny = \n[m]\npath = /x\n",
+      "hosts allow = , ,\n[m]\npath = /x\n",
+      "hosts deny = \t\n[m]\npath = /x\n",
+  };
+  for (size_t i = 0; i < sizeof(empty_hosts) / sizeof(empty_hosts[0]); i++) {
+    EXPECT_EQ_INT(write_conf(empty_hosts[i], &path), 0);
+    const DaemonConf* rejected = daemon_conf_load(path, err, sizeof(err));
+    free(path);
+    EXPECT_NULL(rejected);
+    EXPECT_TRUE(strstr(err, "must list at least one host pattern") != NULL);
+  }
+
+  const char* empty_module_hosts[] = {
+      "[m]\npath = /x\nhosts allow = \n",
+      "[m]\npath = /x\nhosts deny = ,\n",
+  };
+  for (size_t i = 0; i < sizeof(empty_module_hosts) / sizeof(empty_module_hosts[0]); i++) {
+    EXPECT_EQ_INT(write_conf(empty_module_hosts[i], &path), 0);
+    const DaemonConf* rejected = daemon_conf_load(path, err, sizeof(err));
+    free(path);
+    EXPECT_NULL(rejected);
+    EXPECT_TRUE(strstr(err, "must list at least one host pattern") != NULL);
+    /* The diagnostic must name the offending module. */
+    EXPECT_TRUE(strstr(err, "module 'm'") != NULL);
+  }
+
+  /* Per-module host lists APPEND across lines like the global ones.  (A swapped
+   * store_host_list call passed the module name as `replace`, so each line
+   * silently replaced the previous one and only the last survived.) */
+  EXPECT_EQ_INT(write_conf("[m]\npath = /x\n"
+                           "hosts allow = 127.0.0.1\n"
+                           "hosts allow = 10.0.0.0/8\n"
+                           "hosts deny = 192.168.0.1\n"
+                           "hosts deny = 2001:db8::/32\n",
+                           &path),
+                0);
   conf = daemon_conf_load(path, err, sizeof(err));
   free(path);
   EXPECT_NOT_NULL(conf);
-  EXPECT_EQ_INT(conf->global.hosts_allow_count, 0);
+  EXPECT_EQ_INT(conf->modules[0].hosts_allow_count, 2);
+  EXPECT_EQ_STR(conf->modules[0].hosts_allow[0], "127.0.0.1");
+  EXPECT_EQ_STR(conf->modules[0].hosts_allow[1], "10.0.0.0/8");
+  EXPECT_EQ_INT(conf->modules[0].hosts_deny_count, 2);
+  EXPECT_EQ_STR(conf->modules[0].hosts_deny[0], "192.168.0.1");
+  EXPECT_EQ_STR(conf->modules[0].hosts_deny[1], "2001:db8::/32");
   daemon_conf_free(conf);
 }
 

tests/test_file.c

@@ -14,6 +14,7 @@
 #include <stdlib.h>
 #include <string.h>
 #include <sys/stat.h>
+#include <sys/sysmacros.h>
 #include <sys/wait.h>
 #include <time.h>
 #include <unistd.h>
@@ -994,6 +995,94 @@ static void test_inplace_overwrite_truncates_shorter_payload() {
   rmdir(root);
 }
 
+/* B2: --inplace must refuse an existing non-regular destination entry.  A FIFO
+   would block open(O_WRONLY) forever and a device node would be written
+   directly, bypassing the --write-devices/super gate.  Forked with an alarm so
+   a regression is a prompt failure instead of a hung suite. */
+static void test_inplace_refuses_fifo_destination() {
+  const char* root = "test_inplace_fifo_tmp";
+  const char* path = "test_inplace_fifo_tmp/fifo";
+  unlink(path);
+  rmdir(root);
+  EXPECT_EQ_INT(mkdir(root, 0700), 0);
+  EXPECT_EQ_INT(mkfifo(path, 0600), 0);
+
+  pid_t pid = fork();
+  if (pid == 0) {
+    alarm(10);
+    File* f = file_create("fifo");
+    if (!f)
+      _exit(1);
+    const char* content = "payload";
+    f->data->data = malloc(strlen(content));
+    if (!f->data->data)
+      _exit(1);
+    memcpy(f->data->data, content, strlen(content));
+    f->data->size = strlen(content);
+    Config* cfg = config_create();
+    if (!cfg)
+      _exit(1);
+    cfg->inplace = true;
+    bool written = file_save_to_disk(root, f, cfg);
+    file_destroy(f);
+    config_delete(cfg);
+    _exit(written ? 1 : 0); /* must be refused */
+  }
+  int status;
+  waitpid(pid, &status, 0);
+  EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+  struct stat st;
+  EXPECT_EQ_INT(lstat(path, &st), 0);
+  EXPECT_TRUE(S_ISFIFO(st.st_mode)); /* left untouched */
+  unlink(path);
+  rmdir(root);
+}
+
+/* B2: an existing char device must not be written by --inplace.  mknod needs
+   privilege, so a non-root run skips gracefully.  /dev/null's (1:3) rdev makes
+   the negative case harmless if it ever regresses. */
+static void test_inplace_refuses_device_destination() {
+  const char* root = "test_inplace_dev_tmp";
+  const char* path = "test_inplace_dev_tmp/dev";
+  unlink(path);
+  rmdir(root);
+  EXPECT_EQ_INT(mkdir(root, 0700), 0);
+  if (mknod(path, S_IFCHR | 0600, makedev(1, 3)) != 0) {
+    rmdir(root);
+    return; /* no privilege to create a device node: skip */
+  }
+
+  pid_t pid = fork();
+  if (pid == 0) {
+    alarm(10);
+    File* f = file_create("dev");
+    if (!f)
+      _exit(1);
+    const char* content = "payload";
+    f->data->data = malloc(strlen(content));
+    if (!f->data->data)
+      _exit(1);
+    memcpy(f->data->data, content, strlen(content));
+    f->data->size = strlen(content);
+    Config* cfg = config_create();
+    if (!cfg)
+      _exit(1);
+    cfg->inplace = true;
+    bool written = file_save_to_disk(root, f, cfg);
+    file_destroy(f);
+    config_delete(cfg);
+    _exit(written ? 1 : 0); /* must be refused */
+  }
+  int status;
+  waitpid(pid, &status, 0);
+  EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+  struct stat st;
+  EXPECT_EQ_INT(lstat(path, &st), 0);
+  EXPECT_TRUE(S_ISCHR(st.st_mode)); /* still a device, not replaced */
+  unlink(path);
+  rmdir(root);
+}
+
 /* Explicit directory entries (--dirs) create the directory under the receive
    root through the same save funnel, creating parents as needed, and reject
    traversal the same way a file path does. */
@@ -1606,4 +1695,6 @@ void test_file() {
   test_inplace_overwrite_clears_special_mode_bits();
   test_inplace_overwrite_metadata_strips_special_bits();
   test_inplace_overwrite_truncates_shorter_payload();
+  test_inplace_refuses_fifo_destination();
+  test_inplace_refuses_device_destination();
 }

tests/test_file_list.c

@@ -1,5 +1,6 @@
 #include "test_file_list.h"
 #include "file_list.h"
+#include "utils.h"
 #include "test_utils.h"
 #include <stdio.h>
 #include <stdlib.h>
@@ -189,8 +190,35 @@ static void test_deep_paths_are_bounded() {
   free(entry);
 }
 
+/* An over-long list entry must be rejected cleanly instead of being read
+   without a bound (the reader never allocates beyond UTILS_MAX_LINE_LEN). */
+static void test_oversized_entry_rejected() {
+  const char* path = "test_file_list_oversized.txt";
+  FILE* fp = fopen(path, "wb");
+  EXPECT_NOT_NULL(fp);
+  char chunk[4096];
+  memset(chunk, 'a', sizeof(chunk));
+  size_t total = 0;
+  while (total <= UTILS_MAX_LINE_LEN) {
+    EXPECT_EQ_INT((int)fwrite(chunk, 1, sizeof(chunk), fp), (int)sizeof(chunk));
+    total += sizeof(chunk);
+  }
+  EXPECT_EQ_INT(fputc('\n', fp), '\n');
+  fclose(fp);
+
+  char err[160];
+  FileListSet* set = file_list_load(path, false, err, sizeof(err));
+  if (set) {
+    file_list_destroy(set);
+    EXPECT_FAIL("over-long entry was accepted");
+  }
+  EXPECT_TRUE(strstr(err, "exceeds") != NULL);
+  remove(path);
+}
+
 void test_file_list() {
   test_membership_matches_reference();
   test_ancestor_and_descendant_queries();
   test_deep_paths_are_bounded();
-}
+  test_oversized_entry_rejected();
+}

tests/test_glob.c

@@ -1,7 +1,9 @@
 #include "test_glob.h"
 #include "utils.h"
 #include "test_utils.h"
+#include <stdlib.h>
 #include <string.h>
+#include <time.h>
 
 static void test_glob_exact_match() {
   EXPECT_TRUE(glob_match("foo", "foo"));
@@ -76,6 +78,34 @@ static void test_glob_doublestar_mid() {
   EXPECT_FALSE(glob_match("a/**/b", "a/x/bad"));
 }
 
+/* The old backtracking matcher explored an exponential number of paths for a
+ * pattern with many `*` wildcards against a long run that never matches the
+ * trailing literal.  The iterative matcher must stay bounded: 30 `*a` groups
+ * followed by `b` against ten thousand `a`s is a few hundred thousand states,
+ * not 2^30 recursion nodes. */
+static void test_glob_pathological_is_bounded() {
+  char pattern[128];
+  size_t pos = 0;
+  for (int i = 0; i < 30; i++) {
+    pattern[pos++] = '*';
+    pattern[pos++] = 'a';
+  }
+  pattern[pos++] = 'b';
+  pattern[pos] = '\0';
+
+  char* text = malloc(10001);
+  EXPECT_NOT_NULL(text);
+  memset(text, 'a', 10000);
+  text[10000] = '\0';
+
+  clock_t start = clock();
+  EXPECT_FALSE(glob_match(pattern, text));
+  double elapsed = (double)(clock() - start) / CLOCKS_PER_SEC;
+  EXPECT_TRUE(elapsed < 5.0);
+
+  free(text);
+}
+
 void test_glob() {
   test_glob_exact_match();
   test_glob_question_mark();
@@ -91,4 +121,5 @@ void test_glob() {
   test_glob_doublestar_prefix();
   test_glob_doublestar_suffix();
   test_glob_doublestar_mid();
+  test_glob_pathological_is_bounded();
 }

tests/test_server.c

@@ -1,4 +1,5 @@
 #include "test_server.h"
+#include "checksum.h"
 #include "config.h"
 #include "delta.h"
 #include "file.h"
@@ -844,6 +845,258 @@ static void test_special_socket_path_log_escaped() {
   EXPECT_NOT_NULL(strstr(output, "socket not recreated: evil\\#012path"));
 }
 
+/* B1: a client-planted FIFO at the destination must not block the receiver's
+ * incremental-check open.  With the O_NONBLOCK open plus the post-open S_ISREG
+ * gate the FIFO is simply "no existing regular file", so the receiver proceeds
+ * to a full transfer; without O_NONBLOCK the child blocks in openat() and the
+ * alarm(30) kills it. */
+static void test_incremental_check_fifo_destination_does_not_hang() {
+  Config* cfg = config_create();
+  EXPECT_NOT_NULL(cfg);
+  char* root = make_check_root("qffo");
+  EXPECT_NOT_NULL(root);
+  cfg->receive_root_directory = str_dup(root);
+  char path[1024];
+  snprintf(path, sizeof(path), "%s/file.txt", root);
+  EXPECT_EQ_INT(mkfifo(path, 0600), 0);
+
+  int p[2];
+  EXPECT_EQ_INT(socketpair(AF_UNIX, SOCK_STREAM, 0, p), 0);
+  io_set_fds(p[0], p[1]);
+  io_set_bwlimit(0);
+
+  pid_t pid = fork();
+  if (pid == 0) {
+    alarm(30);
+    close(p[1]);
+    io_set_fds(p[0], p[0]);
+    bool skipped = false;
+    File* file = receive_incremental_check(p[0], cfg, &skipped);
+    bool ok = file != NULL && !skipped;
+    file_destroy(file);
+    config_delete(cfg);
+    close(p[0]);
+    _exit(ok ? 0 : 1);
+  } else {
+    close(p[0]);
+    io_set_fds(p[1], p[1]);
+    EXPECT_TRUE(send_str(p[1], "file.txt"));
+    unsigned long long size = 4;
+    long long mtime = 42;
+    long long mtime_nsec = 0;
+    EXPECT_TRUE(send_n_data(p[1], &size, sizeof(size)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime, sizeof(mtime)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime_nsec, sizeof(mtime_nsec)));
+    Status s;
+    EXPECT_TRUE(receive_status(p[1], &s));
+    EXPECT_EQ_INT(s, STATUS_NEXT);
+
+    Data* body = data_create_reserve(4);
+    EXPECT_NOT_NULL(body);
+    body->data = malloc(4);
+    EXPECT_NOT_NULL(body->data);
+    memcpy(body->data, "data", 4);
+    body->size = 4;
+    EXPECT_TRUE(send_data(p[1], body));
+    data_destroy(body);
+
+    int status;
+    waitpid(pid, &status, 0);
+    close(p[1]);
+    config_delete(cfg);
+    unlink(path);
+    rmdir(root);
+    free(root);
+    EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+  }
+}
+
+/* A server-contacting --dry-run with an alternate basis dir must never read or
+   hash the basis file.  An exact (size+mtime+content) basis match would
+   otherwise let a client probe the basis bytes against its own supplied digest
+   (a 1-bit content oracle).  The dry-run decision is metadata-only, so even a
+   byte-identical basis is reported as would-transfer, not a compare-dest skip. */
+static void test_incremental_check_dry_run_basis_does_not_read_content() {
+  Config* cfg = config_create();
+  EXPECT_NOT_NULL(cfg);
+  cfg->dry_run = true;
+  char* root = make_check_root("dryb");
+  EXPECT_NOT_NULL(root);
+  cfg->receive_root_directory = str_dup(root);
+
+  char basis_dir[1024];
+  char basis_path[2048];
+  snprintf(basis_dir, sizeof(basis_dir), "%s/basis", root);
+  EXPECT_EQ_INT(mkdir(basis_dir, 0700), 0);
+  const char* content = "basis content that matches\n";
+  write_check_file(basis_dir, "file.txt", content);
+  snprintf(basis_path, sizeof(basis_path), "%s/file.txt", basis_dir);
+  struct stat bst;
+  EXPECT_EQ_INT(stat(basis_path, &bst), 0);
+  EXPECT_EQ_INT(config_basis_append(cfg, BASIS_DEST_COMPARE, "basis"), 0);
+
+  /* The (correct) source digest for the basis bytes: an unfixed dry-run would
+     read+hash the basis and treat this as an exact compare-dest hit. */
+  uint8_t digest[CHECKSUM_MAX_DIGEST_LEN];
+  size_t digest_len = 0;
+  EXPECT_TRUE(checksum_digest((ChecksumAlgo)cfg->checksum_algo, cfg->checksum_seed, content,
+                              strlen(content), digest, sizeof(digest), &digest_len));
+
+  int p[2];
+  EXPECT_EQ_INT(socketpair(AF_UNIX, SOCK_STREAM, 0, p), 0);
+  io_set_fds(p[0], p[1]);
+  io_set_bwlimit(0);
+
+  pid_t pid = fork();
+  if (pid == 0) {
+    alarm(30);
+    close(p[1]);
+    io_set_fds(p[0], p[0]);
+    bool skipped = false;
+    bool would_transfer = false;
+    File* file = receive_incremental_check_ex(p[0], cfg, &skipped, &would_transfer);
+    bool ok = file == NULL && !skipped && would_transfer;
+    file_destroy(file);
+    config_delete(cfg);
+    close(p[0]);
+    _exit(ok ? 0 : 1);
+  } else {
+    close(p[0]);
+    io_set_fds(p[1], p[1]);
+    EXPECT_TRUE(send_str(p[1], "file.txt"));
+    unsigned long long size = (unsigned long long)bst.st_size;
+    long long mtime = (long long)bst.st_mtime;
+    long long mtime_nsec = 0;
+#ifdef __linux__
+    mtime_nsec = (long long)bst.st_mtim.tv_nsec;
+#endif
+    EXPECT_TRUE(send_n_data(p[1], &size, sizeof(size)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime, sizeof(mtime)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime_nsec, sizeof(mtime_nsec)));
+    uint8_t wire_len = (uint8_t)digest_len;
+    EXPECT_TRUE(send_n_data(p[1], &wire_len, sizeof(wire_len)));
+    EXPECT_TRUE(send_n_data(p[1], digest, digest_len));
+    Status s;
+    EXPECT_TRUE(receive_status(p[1], &s));
+    /* A skip here would mean the receiver read+hashed the basis file. */
+    EXPECT_EQ_INT(s, STATUS_DRY_RUN_TRANSFER);
+
+    int status;
+    waitpid(pid, &status, 0);
+    close(p[1]);
+    config_delete(cfg);
+    /* The dry-run must not have materialized anything in the receive root. */
+    char dest_path[2048];
+    snprintf(dest_path, sizeof(dest_path), "%s/file.txt", root);
+    EXPECT_FALSE(file_path_exists_secure(dest_path));
+    unlink(basis_path);
+    rmdir(basis_dir);
+    rmdir(root);
+    free(root);
+    EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+  }
+}
+
+/* B1: a FIFO planted in a --link-dest basis directory must not block
+ * basis_open_regular() either; the basis match is simply declined. */
+static void test_incremental_check_basis_fifo_does_not_hang() {
+  Config* cfg = config_create();
+  EXPECT_NOT_NULL(cfg);
+  char* root = make_check_root("qbfi");
+  EXPECT_NOT_NULL(root);
+  cfg->receive_root_directory = str_dup(root);
+  char basis_dir[1024];
+  char basis_path[2048];
+  snprintf(basis_dir, sizeof(basis_dir), "%s/basis", root);
+  EXPECT_EQ_INT(mkdir(basis_dir, 0700), 0);
+  snprintf(basis_path, sizeof(basis_path), "%s/file.txt", basis_dir);
+  EXPECT_EQ_INT(mkfifo(basis_path, 0600), 0);
+  EXPECT_EQ_INT(config_basis_append(cfg, BASIS_DEST_LINK, "basis"), 0);
+
+  int p[2];
+  EXPECT_EQ_INT(socketpair(AF_UNIX, SOCK_STREAM, 0, p), 0);
+  io_set_fds(p[0], p[1]);
+  io_set_bwlimit(0);
+
+  pid_t pid = fork();
+  if (pid == 0) {
+    alarm(30);
+    close(p[1]);
+    io_set_fds(p[0], p[0]);
+    bool skipped = false;
+    File* file = receive_incremental_check(p[0], cfg, &skipped);
+    bool ok = file != NULL && !skipped;
+    file_destroy(file);
+    config_delete(cfg);
+    close(p[0]);
+    _exit(ok ? 0 : 1);
+  } else {
+    close(p[0]);
+    io_set_fds(p[1], p[1]);
+    EXPECT_TRUE(send_str(p[1], "file.txt"));
+    unsigned long long size = 4;
+    long long mtime = 42;
+    long long mtime_nsec = 0;
+    EXPECT_TRUE(send_n_data(p[1], &size, sizeof(size)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime, sizeof(mtime)));
+    EXPECT_TRUE(send_n_data(p[1], &mtime_nsec, sizeof(mtime_nsec)));
+    /* config_has_basis() makes the request carry the source digest. */
+    uint8_t wire_len = 8;
+    uint8_t digest[8] = {0};
+    EXPECT_TRUE(send_n_data(p[1], &wire_len, sizeof(wire_len)));
+    EXPECT_TRUE(send_n_data(p[1], digest, sizeof(digest)));
+    Status s;
+    EXPECT_TRUE(receive_status(p[1], &s));
+    EXPECT_EQ_INT(s, STATUS_NEXT);
+
+    Data* body = data_create_reserve(4);
+    EXPECT_NOT_NULL(body);
+    body->data = malloc(4);
+    EXPECT_NOT_NULL(body->data);
+    memcpy(body->data, "data", 4);
+    body->size = 4;
+    EXPECT_TRUE(send_data(p[1], body));
+    data_destroy(body);
+
+    int status;
+    waitpid(pid, &status, 0);
+    close(p[1]);
+    config_delete(cfg);
+    unlink(basis_path);
+    rmdir(basis_dir);
+    rmdir(root);
+    free(root);
+    EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
+  }
+}
+
+/* B5: the aggregate entry count across the three manifest sections is capped at
+ * MAX_MANIFEST_ENTRIES, and a section that would push the total over the cap is
+ * rejected before its entries are read (so a tiny first section followed by a
+ * huge claimed second section fails fast). */
+static void test_receive_manifest_total_entry_cap() {
+  Config* cfg = config_create();
+  EXPECT_NOT_NULL(cfg);
+  cfg->receive_root_directory = str_dup("/tmp/dst");
+  int p[2];
+  EXPECT_EQ_INT(socketpair(AF_UNIX, SOCK_STREAM, 0, p), 0);
+  io_set_fds(p[0], p[1]);
+
+  EXPECT_TRUE(send_int(p[1], 1));
+  EXPECT_TRUE(send_str(p[1], "keep.txt"));
+  /* The second section alone is within its per-section cap, but 1 + it exceeds
+     the cross-section cap; the receiver must reject at the count. */
+  EXPECT_TRUE(send_int(p[1], MAX_MANIFEST_ENTRIES));
+  EXPECT_NULL(receive_manifest_entries(p[0]));
+  Status status;
+  EXPECT_TRUE(receive_status(p[1], &status));
+  EXPECT_EQ_INT(status, STATUS_ERROR);
+
+  close(p[0]);
+  close(p[1]);
+  config_delete(cfg);
+}
+
 void test_server() {
   test_special_socket_path_log_escaped();
   if (!is_running_under_valgrind()) {
@@ -856,6 +1109,10 @@ void test_server() {
     test_incremental_check_size_mismatch_full_transfer();
     test_incremental_check_dry_run_reports_transfer_without_writing();
     test_incremental_check_delta_oversize_reports_failure();
+    test_incremental_check_fifo_destination_does_not_hang();
+    test_incremental_check_basis_fifo_does_not_hang();
+    test_incremental_check_dry_run_basis_does_not_read_content();
+    test_receive_manifest_total_entry_cap();
     test_late_manifest_abort_frees_keepset();
     test_late_manifest_eof_frees_keepset();
     test_late_second_manifest_frees_both();

tests/test_server_cli.c

@@ -32,6 +32,35 @@ static void test_server_cli_defaults() {
   EXPECT_FALSE(opts.allow_delete);
   EXPECT_FALSE(opts.allow_unauthenticated);
   EXPECT_FALSE(opts.no_super);
+  EXPECT_FALSE(opts.allow_super);
+  server_cli_options_free(&opts);
+}
+
+/* C3: --allow-super is the locally-launched standalone TCP opt-in for a
+ * privileged receiver; it never combines with --no-super, is refused with
+ * --stdio (whose client-composed remote argv must not defeat the default), and
+ * daemon modules use their own per-module `client owner = yes` opt-in instead. */
+static void test_server_cli_allow_super() {
+  const char* args[] = {"fastsync-server", "--allow-super", "--destination-root", "/srv"};
+  ServerCliOptions opts;
+  EXPECT_EQ_INT(parse_ok(args, 4, &opts), 0);
+  EXPECT_TRUE(opts.allow_super);
+  server_cli_options_free(&opts);
+
+  char err[256];
+  const char* a1[] = {"s", "--allow-super", "--no-super"};
+  EXPECT_EQ_INT(server_cli_parse(3, (char**)a1, &opts, err, sizeof(err)), -1);
+  EXPECT_TRUE(strstr(err, "mutually exclusive") != NULL);
+
+  const char* a2[] = {"s", "--daemon", "--config=/tmp/x.conf", "--allow-super"};
+  EXPECT_EQ_INT(server_cli_parse(4, (char**)a2, &opts, err, sizeof(err)), -1);
+  EXPECT_TRUE(strstr(err, "client owner") != NULL);
+
+  /* The SSH/--stdio receiver argv is composed by the client, so --allow-super
+   * must be rejected there and the C3 secure default stays in force. */
+  const char* a3[] = {"s", "--stdio", "--allow-super", "--destination-root", "/srv"};
+  EXPECT_EQ_INT(server_cli_parse(5, (char**)a3, &opts, err, sizeof(err)), -1);
+  EXPECT_TRUE(strstr(err, "--stdio") != NULL);
   server_cli_options_free(&opts);
 }
 
@@ -224,5 +253,6 @@ void test_server_cli() {
   test_server_cli_password_and_early_input();
   test_server_cli_password_requires_daemon();
   test_server_cli_no_super();
+  test_server_cli_allow_super();
   test_server_cli_help();
 }

tests/test_shared_utils.c

@@ -519,9 +519,38 @@ static void test_path_index_semantics() {
   path_index_free(&empty);
 }
 
+/* utils_getdelim_bounded must return normal short lines unchanged and refuse an
+ * over-long record with EFBIG rather than allocating without bound. */
+static void test_getdelim_bounded() {
+  FILE* fp = tmpfile();
+  EXPECT_NOT_NULL(fp);
+  const char* short_line = "short\n";
+  EXPECT_EQ_INT((int)fwrite(short_line, 1, strlen(short_line), fp), (int)strlen(short_line));
+  char big[32];
+  memset(big, 'x', 20);
+  big[20] = '\n';
+  EXPECT_EQ_INT((int)fwrite(big, 1, 21, fp), 21);
+  rewind(fp);
+
+  char* line = NULL;
+  size_t cap = 0;
+  ssize_t n = utils_getdelim_bounded(fp, &line, &cap, '\n', 64);
+  EXPECT_EQ_INT((int)n, 6);
+  EXPECT_EQ_STR(line, "short\n");
+
+  errno = 0;
+  n = utils_getdelim_bounded(fp, &line, &cap, '\n', 10);
+  EXPECT_EQ_INT((int)n, -1);
+  EXPECT_EQ_INT(errno, EFBIG);
+
+  free(line);
+  fclose(fp);
+}
+
 void test_shared_utils() {
   test_path_index_bounded();
   test_path_index_semantics();
+  test_getdelim_bounded();
   test_walker_removes_extras_keeps_manifest_and_protected();
   test_walker_keeps_nested_manifest_dirs();
   test_walker_max_delete_exceeded_deletes_nothing();

tests/test_transport_ssh.c

@@ -66,16 +66,18 @@ static void test_ssh_remote_command_argument_modes() {
   free(command);
 }
 
-/* The build for a single-word argv is [prog, six -o args, user, command]. */
+/* The build for a single-word argv is [prog, six -o args, "--", user, command]. */
 
 static void test_ssh_build_client_argv_default_is_ssh() {
   char** argv = ssh_build_client_argv(NULL, 0, "u@h", "'srv' --stdio");
   EXPECT_NOT_NULL(argv);
   EXPECT_EQ_STR(argv[0], "ssh");
   EXPECT_EQ_STR(argv[1], "-o");
-  EXPECT_EQ_STR(argv[7], "u@h");
-  EXPECT_EQ_STR(argv[8], "'srv' --stdio");
-  EXPECT_NULL(argv[9]);
+  /* The "--" end-of-options marker precedes the destination token. */
+  EXPECT_EQ_STR(argv[7], "--");
+  EXPECT_EQ_STR(argv[8], "u@h");
+  EXPECT_EQ_STR(argv[9], "'srv' --stdio");
+  EXPECT_NULL(argv[10]);
   ssh_free_client_argv(argv);
 }
 
@@ -84,7 +86,7 @@ static void test_ssh_build_client_argv_uses_custom_rsh() {
   char** argv = ssh_build_client_argv("myrsh", 0, "u@h", "rc");
   EXPECT_NOT_NULL(argv);
   EXPECT_EQ_STR(argv[0], "myrsh");
-  EXPECT_NULL(argv[9]);
+  EXPECT_NULL(argv[10]);
   ssh_free_client_argv(argv);
 }
 
@@ -96,21 +98,37 @@ static void test_ssh_build_client_argv_whitespace_command_and_port() {
   EXPECT_EQ_STR(argv[0], "ssh");
   EXPECT_EQ_STR(argv[1], "-p");
   EXPECT_EQ_STR(argv[2], "2222");
-  EXPECT_NULL(argv[11]);
+  EXPECT_NULL(argv[12]);
   ssh_free_client_argv(argv);
 
   argv = ssh_build_client_argv("ssh", 2222, "u@h", "rc");
   EXPECT_NOT_NULL(argv);
   EXPECT_EQ_STR(argv[0], "ssh");
-  /* Flat [prog, -o x6, -p, port, user, command]. */
+  /* Flat [prog, -o x6, -p, port, "--", user, command]. */
   EXPECT_EQ_STR(argv[7], "-p");
   EXPECT_EQ_STR(argv[8], "2222");
-  EXPECT_EQ_STR(argv[9], "u@h");
-  EXPECT_EQ_STR(argv[10], "rc");
-  EXPECT_NULL(argv[11]);
+  EXPECT_EQ_STR(argv[9], "--");
+  EXPECT_EQ_STR(argv[10], "u@h");
+  EXPECT_EQ_STR(argv[11], "rc");
+  EXPECT_NULL(argv[12]);
   ssh_free_client_argv(argv);
 }
 
+/* C1: a destination host/user beginning with '-' would be parsed by ssh as an
+ * option (argument injection: -oProxyCommand=...), and an empty host is never
+ * valid.  These are refused before any child is forked, so no Client is
+ * returned and no command can run. */
+static void test_ssh_connect_rejects_option_host() {
+  /* cppcheck-suppress constVariablePointer */
+  Client* client = client_connect_ssh("-oProxyCommand=touch /tmp/pwned:/remote", 22, NULL, false,
+                                      NULL, false, NULL, 0);
+  EXPECT_NULL(client);
+  client = client_connect_ssh("-evil:/remote", 22, NULL, false, NULL, false, NULL, 0);
+  EXPECT_NULL(client);
+  client = client_connect_ssh("user@:/remote", 22, NULL, false, NULL, false, NULL, 0);
+  EXPECT_NULL(client);
+}
+
 /* --remote-option=OPT appends OPT to the remote command line after " --stdio",
  * each escaped as its own single-quoted shell word.  Metacharacters that could
  * break out of the quoting are neutralized (never injected), matching the
@@ -162,6 +180,7 @@ void test_transport_ssh() {
   test_ssh_connect_invalid_dest_empty();
   test_ssh_connect_malformed();
   test_ssh_connect_unreachable();
+  test_ssh_connect_rejects_option_host();
   test_ssh_remote_command_argument_modes();
   test_ssh_build_client_argv_default_is_ssh();
   test_ssh_build_client_argv_uses_custom_rsh();

tests/test_transport_tls.c

@@ -24,6 +24,17 @@ static void test_server_create_tls_without_certs() {
 #ifdef SSL_OP_NO_RENEGOTIATION
   EXPECT_TRUE((SSL_CTX_get_options(ctx) & SSL_OP_NO_RENEGOTIATION) != 0);
 #endif
+  /* C5: the server's preference order decides the cipher and the TLS 1.2 list is
+   * AEAD-only (no CBC/RC4/3DES legacy suites). */
+  EXPECT_TRUE((SSL_CTX_get_options(ctx) & SSL_OP_CIPHER_SERVER_PREFERENCE) != 0);
+  STACK_OF(SSL_CIPHER)* ciphers = SSL_CTX_get_ciphers(ctx);
+  EXPECT_NOT_NULL(ciphers);
+  for (int i = 0; i < sk_SSL_CIPHER_num(ciphers); i++) {
+    const char* name = SSL_CIPHER_get_name(sk_SSL_CIPHER_value(ciphers, i));
+    EXPECT_TRUE(name != NULL && strstr(name, "CBC") == NULL);
+    EXPECT_TRUE(name != NULL && strstr(name, "RC4") == NULL);
+    EXPECT_TRUE(name != NULL && strstr(name, "3DES") == NULL);
+  }
   server_delete(&s);
   EXPECT_NULL(s);
 }

tests/test_xattr.c

@@ -17,7 +17,7 @@
 
 static void run_recv_helper(int fd) {
   int ok = 0;
-  FileXattrList* list = xattr_receive(fd, &ok);
+  FileXattrList* list = xattr_receive(fd, &ok, false);
   if (!ok)
     _exit(1);
   if (!list) {
@@ -64,7 +64,7 @@ static void test_xattr_wire_roundtrip() {
 
 static void run_recv_must_fail(int fd) {
   int ok = 0;
-  FileXattrList* list = xattr_receive(fd, &ok);
+  FileXattrList* list = xattr_receive(fd, &ok, false);
   /* A NULL list with ok==0 is the expected rejection. */
   if (ok == 0 && list == NULL)
     _exit(0);
@@ -148,15 +148,64 @@ static void test_xattr_count_bound() {
 /* The captured list on a plain file reflects only whitelisted namespaces
  * (Linux only; skipped when the filesystem has no xattr support). */
 static void test_xattr_capture_and_appliable() {
-  EXPECT_FALSE(xattr_name_appliable(NULL));
-  EXPECT_FALSE(xattr_name_appliable(""));
-  EXPECT_FALSE(xattr_name_appliable("security.selinux"));
-  EXPECT_FALSE(xattr_name_appliable("trusted.blob"));
-  EXPECT_TRUE(xattr_name_appliable("user.foo"));
+  EXPECT_FALSE(xattr_name_appliable(NULL, false));
+  EXPECT_FALSE(xattr_name_appliable("", false));
+  EXPECT_FALSE(xattr_name_appliable("security.selinux", false));
+  EXPECT_FALSE(xattr_name_appliable("trusted.blob", false));
+  EXPECT_TRUE(xattr_name_appliable("user.foo", false));
+  EXPECT_TRUE(xattr_name_appliable("user.foo", true));
   /* The reserved fake-super key is receiver-only and never forwarded/applied. */
-  EXPECT_FALSE(xattr_name_appliable("user.fastsync.stat"));
-  EXPECT_TRUE(xattr_name_appliable("system.posix_acl_access"));
-  EXPECT_TRUE(xattr_name_appliable("system.posix_acl_default"));
+  EXPECT_FALSE(xattr_name_appliable("user.fastsync.stat", false));
+  EXPECT_FALSE(xattr_name_appliable("user.fastsync.stat", true));
+  /* B4: the ACL names require --acls; -X alone must not authorize them. */
+  EXPECT_FALSE(xattr_name_appliable("system.posix_acl_access", false));
+  EXPECT_FALSE(xattr_name_appliable("system.posix_acl_default", false));
+  EXPECT_TRUE(xattr_name_appliable("system.posix_acl_access", true));
+  EXPECT_TRUE(xattr_name_appliable("system.posix_acl_default", true));
+}
+
+/* B4: a `-X`-only receiver (preserve_acls false) must NOT apply an incoming
+ * ACL xattr, while a user.* attribute in the same block still survives.  The
+ * ACL entry is dropped, not applied (and the -X transfer is not failed). */
+static void run_recv_drops_acl_keeps_user(int fd) {
+  int ok = 0;
+  FileXattrList* list = xattr_receive(fd, &ok, false);
+  if (!ok || list == NULL)
+    _exit(1);
+  bool saw_user = false;
+  for (int i = 0; i < list->count; i++) {
+    if (strcmp(list->items[i].name, "system.posix_acl_access") == 0)
+      _exit(1); /* ACL must have been dropped */
+    if (strcmp(list->items[i].name, "user.keep") == 0)
+      saw_user = true;
+  }
+  xattr_list_free(list);
+  _exit(saw_user ? 0 : 1);
+}
+
+static void test_xattr_receive_drops_acl_without_preserve_acls() {
+  int p[2];
+  EXPECT_EQ_INT(pipe(p), 0);
+  io_set_fds(p[0], p[1]);
+  io_set_bwlimit(0);
+  pid_t pid = fork();
+  if (pid == 0) {
+    close(p[1]);
+    io_set_fds(p[0], p[0]);
+    run_recv_drops_acl_keeps_user(p[0]);
+  }
+  close(p[0]);
+  io_set_fds(p[1], p[1]);
+  FileXattrList* list = xattr_list_new();
+  EXPECT_NOT_NULL(list);
+  EXPECT_TRUE(xattr_list_append(list, "system.posix_acl_access", "\x02\x00\x00\x00", 4));
+  EXPECT_TRUE(xattr_list_append(list, "user.keep", "yes", 3));
+  xattr_send(p[1], list);
+  xattr_list_free(list);
+  int status;
+  waitpid(pid, &status, 0);
+  close(p[1]);
+  EXPECT_TRUE(WIFEXITED(status) && WEXITSTATUS(status) == 0);
 }
 
 /* MINOR-2: a --link-dest / -H copy fallback (linkat refused) must still apply
@@ -224,6 +273,80 @@ static void test_link_copy_fallback_preserves_xattrs() {
   rmdir(basis_dir);
 }
 
+/* Capture must honor --acls: xattr_capture_path(path, false) (plain -X) must
+ * never return the POSIX ACL names, while xattr_capture_path(path, true) (-A)
+ * does; user.* is captured either way.  This is the capture-side counterpart of
+ * the receiver's --acls gate and must not depend on the caller having checked
+ * the flag.  Guarded on filesystem/ACL support. */
+static void test_xattr_capture_filters_acls() {
+  const char* path = "test_xattr_capture_acls.txt";
+  unlink(path);
+  int fd = open(path, O_WRONLY | O_CREAT | O_TRUNC, 0600);
+  if (fd < 0)
+    return;
+  bool has_xattr = setxattr(path, "user.fastsync.xprobe", "p", 1, 0) == 0;
+  if (has_xattr)
+    removexattr(path, "user.fastsync.xprobe");
+  if (!has_xattr) {
+    close(fd);
+    unlink(path);
+    return; /* filesystem without xattr support */
+  }
+  if (setxattr(path, "user.keep", "yes", 3, 0) != 0) {
+    close(fd);
+    unlink(path);
+    return;
+  }
+
+  /* Synthesize a valid non-trivial POSIX access ACL blob (little-endian):
+     version 2 followed by USER_OBJ/USER/GROUP_OBJ/MASK/OTHER entries. */
+  uint32_t acl_uid = geteuid() == 0 ? 65534u : (uint32_t)geteuid();
+  unsigned char blob[4 + 5 * 8];
+  uint32_t version = 2;
+  memcpy(blob, &version, 4);
+  const uint16_t tags[5] = {0x01, 0x02, 0x04, 0x10, 0x20}; /* OBJ/USER/GROUP/MASK/OTHER */
+  const uint16_t perms[5] = {0x04, 0x04, 0x04, 0x04, 0x00};
+  const uint32_t ids[5] = {0xFFFFFFFFu, acl_uid, 0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu};
+  size_t off = 4;
+  for (int i = 0; i < 5; i++) {
+    memcpy(blob + off, &tags[i], sizeof(tags[i]));
+    off += sizeof(tags[i]);
+    memcpy(blob + off, &perms[i], sizeof(perms[i]));
+    off += sizeof(perms[i]);
+    memcpy(blob + off, &ids[i], sizeof(ids[i]));
+    off += sizeof(ids[i]);
+  }
+  if (setxattr(path, "system.posix_acl_access", blob, off, 0) != 0) {
+    close(fd);
+    unlink(path);
+    return; /* no unprivileged ACL support: skip silently */
+  }
+  close(fd);
+
+  FileXattrList* plain = xattr_capture_path(path, false);
+  FileXattrList* with_acls = xattr_capture_path(path, true);
+  bool plain_user = false, plain_acl = false, acl_user = false, acl_acl = false;
+  for (int i = 0; plain && i < plain->count; i++) {
+    if (strcmp(plain->items[i].name, "user.keep") == 0)
+      plain_user = true;
+    if (strcmp(plain->items[i].name, "system.posix_acl_access") == 0)
+      plain_acl = true;
+  }
+  for (int i = 0; with_acls && i < with_acls->count; i++) {
+    if (strcmp(with_acls->items[i].name, "user.keep") == 0)
+      acl_user = true;
+    if (strcmp(with_acls->items[i].name, "system.posix_acl_access") == 0)
+      acl_acl = true;
+  }
+  EXPECT_TRUE(plain_user);
+  EXPECT_FALSE(plain_acl);
+  EXPECT_TRUE(acl_user);
+  EXPECT_TRUE(acl_acl);
+  xattr_list_free(plain);
+  xattr_list_free(with_acls);
+  unlink(path);
+}
+
 /* --fake-super replay: fake_super_store_fd records the source stat into the
  * reserved xattr, and fake_super_restore_fd re-applies mode/mtime (and owner,
  * when the process may) fd-relative.  Restore must also be a safe no-op with no
@@ -360,6 +483,8 @@ void test_xattr() {
   test_xattr_reject_oversized_value();
   test_xattr_count_bound();
   test_xattr_capture_and_appliable();
+  test_xattr_capture_filters_acls();
+  test_xattr_receive_drops_acl_without_preserve_acls();
   test_link_copy_fallback_preserves_xattrs();
   test_fake_super_restore();
   test_fake_super_owner_gate();