xeiaso.net
Proof of concept mechanical port of ircnet/ircd to Rust as part of a bit about C being insecure for network services
1.6 kB
28 lines
1---
2name: p7-l1-shared-global-race
3description: "P7 L1 differential tests that touch shared C globals (me/cref_me, local[], highest_fd, timeofday) must serialize on a mutex — cargo runs"
4metadata:
5 node_type: memory
6 type: feedback
7 originSessionId: 7cceaba5-1974-4889-94b9-afa6039d571f
8---
9
10P7 (I/O core) ports touch process-wide C globals (`me`/`cref_me`, and coming up
11`local[]`, `highest_fd`, `timeofday`, `FdAry`). An L1 differential test that zeroes
12then reads such a global races against cargo's other parallel `#[test]` threads in
13the same binary — one test zeroes `me` while another reads it → flaky wrong values
14(seen in P7a `bsd_diff.rs`: `cref_me.sendB` read 0 instead of the written value;
15the failing case passed when run alone via a filter, which is the tell).
16
17**Why:** unlike heap clients passed by pointer (per-test, no sharing), a `static mut`
18global is one instance shared by all tests in the binary. This is the L1 analogue of
19the L2 `PORT_LOCK`.
20
21**How to apply:** in any L1 test file that mutates+reads a shared C global, add a
22file-level `static LOCK: std::sync::Mutex<()> = Mutex::new(())` and have the per-test
23`setup()` return the held guard (`LOCK.lock().unwrap_or_else(|e| e.into_inner())`);
24bind it `let _guard = setup();` so it's held for the whole body. See
25`ircd-testkit/tests/bsd_diff.rs`. Diagnose a suspected race by re-running the single
26failing test with a name filter — if it passes alone but fails in the suite, it's this.
27
28Related: [[l1-cref-static-symbol-limit]], [[piauth-state]] (P7 is the linear next phase).