glm-5.3-flash
0ae53d2d27
FWD-21, found minutes into the first ws_session_opseq campaign: one
feed that drains a complete chunk and then hits an oversize header
(or the pending cap) in the same call returned
Walk::Oversize/PendingOverCap and silently discarded the already
drained pieces — the walk's own consumption-accounting contract
('every parsed byte is either emitted or retained', wave-1
invariant 6) was violated by the crate. Production could not reach
the shape (the mux's per-write invariant caps pending at one chunk +
header), so impact was nil, but WriteChunkWalk is public API and the
wave-1 split_walk equality check was skipped on exactly this path.
- Walk::Oversize(Vec<Vec<u8>>) / Walk::PendingOverCap(Vec<Vec<u8>>)
carry the pre-failure pieces (user-approved semver-relevant API
change)
- run_write_pump emits the carried pieces before fail_write_pump, so
a 1011 close never swallows a legal chunk the mux already counted
as written
- the ws_chunk_framing harness's split-drive failure arm now
reconciles the parsed-before-failure bytes across the whole and
split drives; a regression seed covers the drain-then-fail shape
Verification: main crate 441 default / 520 all-features tests, fuzz
shared replay 21/21 (default + no-default-features), cargo fuzz build
clean, clippy+fmt clean both crates.
alkhttp fuzzing
cargo-fuzz targets for the HTTP surface's hand-rolled parse paths. The
design and operating rules live in docs/plans/fuzzing.md (adopted
from alkcall's docs/research/fuzzing.md); this README is the
operational cheat-sheet.
Layout
fuzz_targets/— nightly-onlyfuzz_target!binaries (thin wrappers).shared/— stable-toolchain library holding the invariant logic; the corpus replay tests run here on plaincargo test.corpus/<target>/— committed seeds (regenerate withpython3 fuzz/gen_fuzz_seeds.py).artifacts/— gitignored crash/oom/timeout artifacts + campaign logs.
Targets
| Target | Drives |
|---|---|
sse_parser |
SseParser::feed + the FWD-17 envelope mapping (src/adapters/forward.rs) — the SSE frame parser every subscription forwards through |
ws_chunk_framing |
WriteChunkWalk (the write pump's chunk walk) (src/websocket/byte_adapter.rs) |
spec_import |
OpenAPISpec::from_str/from_yaml/from_json, the YAML normalization boundary, CompiledInputSchema compile/harden/validate, FromMCP::build_spec (src/adapters/openapi_spec.rs, input_validation.rs, from_mcp/mod.rs) |
Running a campaign — always detached
Agent sessions must never run fuzzing in the foreground (an OOM in a target can take down the session host; see docs/plans/fuzzing.md §2). Use the detached runner:
fuzz/run-detached.sh sse_parser
# poll:
tail -n 50 fuzz/artifacts/sse_parser-*.log
ls fuzz/artifacts/sse_parser/
pgrep -f "cargo fuzz run sse_parser"
FUZZ_RUNTIME_SECS=1800 fuzz/run-detached.sh sse_parser for a longer
campaign. The runner pins -fork=1 -rss_limit_mb=2048 -malloc_limit_mb=2048 -timeout=25 and detaches via setsid + nohup.
Corpus replay (the standing fuzz gate)
cargo test --manifest-path fuzz/shared/Cargo.toml
Replays every committed seed through the same invariant functions the fuzzer runs, on stable, no nightly needed. Part of the release verification checklist (AGENTS.md).
Toolchain notes
rust-toolchain.toml pins nightly for this subtree only (llvm-tools
required by cargo-fuzz); the crate itself stays on stable at MSRV 1.88.
cargo fuzz build must be run with the fuzz dir as CWD or trust the
toolchain file (rustup resolves per directory).
Run campaigns before releases and after touching src/adapters/forward.rs
(the SSE parser), src/websocket/byte_adapter.rs (the WS chunk walk),
or the import adapters (src/adapters/openapi_spec.rs,
input_validation.rs, from_mcp/).