docs(lane): reductions were never a bisect, and the onset is not a source boundary

Three corrections to what this ledger still asserted.

  - It called the reductions a "bisect". They were not: no run isolated
    a variable, because the subsets and the sweeps executed different
    Cargo compilations.
  - It said those compilations were "byte-different". Withdrawn --- the
    bytes a historical run executed are not knowable now, since target
    dirs have been overwritten and a hash computed today is the current
    occupant's. What IS known is the differing Cargo suffixes, i.e.
    differing metadata hashes, which is enough to void the comparison
    and is all that is claimed.
  - The onset was described as a bisectable window. It is not. Reflog
    and commit times put HEAD at 7599661 during the last green ---
    3c06176 landed 40s after that run finished --- and at 724b785
    during the first red, since 5174f73 landed 08:45:41 after that run
    ended 08:42:01. Cleanliness was captured at neither endpoint.
    Decisively, 72da24a is an ANCESTOR of the passing 7599661 yet fails
    today, and no source-monotonic cause produces that. Reproduce clean
    endpoints before considering a bisect.

Also corrects the onset counts: per test copy the 17 sweep-crdt logs are
13 both-copies-ok, 1 where neither copy executed (the stage died
compiling pmacs with error[E0308]), and 3 both-failed. "14 passes" and
"the earlier reds failed on unrelated tests" were both wrong --- one of
those runs never reached the test at all.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016bqGA6s9tTUFzYpbeW3tai

docs/active-work.md

@@ -494,8 +494,8 @@ from #171 and #215.
     enumerated with its own log digest in the teardown lane's
     `docs/probe-sigint-evidence.md`.
   - **The defect now has its own lane:
-    `gpu-probe-sigint-teardown`** (pushed; framing revision 3 at
+    `gpu-probe-sigint-teardown`** (pushed; framing revision 5 at
-    `docs/gpu-probe-sigint-framing.md`, run provenance at
+    `docs/gpu-probe-sigint-framing.md` (revision 5), run provenance at
     `docs/probe-sigint-evidence.md`). **§5b is held behind it.** That
     lane's framing supersedes every diagnostic claim below; the entries
     here are kept only as the record of how it was found.
@@ -559,15 +559,19 @@ from #171 and #215.
     - Note the test exists in **two** binaries: `tests/gpu_initial_
       target_acceptance.rs` includes it as a module, so a reproducing
       sweep fails it twice, at log lines 3083 and 3117.
-  - **What the bisect established, and what it did NOT.** Green in
+  - **What the reductions established, and what they did NOT** — they
+    were never a bisect, since no run isolated a variable. Green in
     every smaller context tried — the test alone (x3, 0.15s against its
     5s deadline); its whole 15-test suite; a workspace-wide run
     FILTERED to just this test; the lib binary (2145 tests) then the
     suite; the three GPU suites in sweep order — and red in every full
     workspace sweep, seven of them across two trees.
     **No cumulative cause follows from that.** The subset runs and the
-    sweeps executed byte-different binaries, so the reductions never
+    sweeps executed **different Cargo compilations** — differing
-    made the comparison they appeared to make. What is established is
+    suffixes, hence differing metadata hashes — so the reductions never
+    made the comparison they appeared to make. (Calling them
+    "byte-different" is withdrawn: the bytes a historical run executed
+    are not knowable now, only the suffixes.) What is established is
     only: reproducible in the full sweep, not reproduced in any subset
     attempted so far.
   - **`/tmp` is a 30G tmpfs holding 21G**, almost all stale