Per review of aceea70e: a dirty checkout is refused (HEAD would not
describe the code that runs, and a dirty-to-dirty edit would pass the
end-of-run check). On a clean tree HEAD pins the imported helpers too.
plan.json and result.json carry schema=2; schema-1 receipts (bare list,
e.g. evidence 70f6617) are documented. Adds a dirty-refusal test.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- test_refined_dfg::test_source_pin runs its mismatch negative controls
first (now including window mode) so missing history cannot skip them.
- test_re100: renamed to test_manifest_matches_its_pinned_commit, plus an
explicit record that today's src differs from the pin.
- Structural manifest checks run before any history-dependent check.
- Real tamper test: a flipped digest raises via verify_manifest, and a
path absent at the commit raises ValueError.
- A blobless clone (tree present, blob missing) is reported as unavailable
history, not as a missing path.
- All history skips go through history_unavailable(): fatal when
CELERIS_REQUIRE_HISTORY=1.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The a135252 manifest is historical qualification: it pins the committed
corrected tree at a135252. The guard is correct; the four tests used the
current working tree as the 'pinned' fixture, so 87c3a31's jets.py change
broke them. Fixtures are now materialized from a135252 with git; the
corrected acceptance, baseline rejection, four tamper rejections and the
reversed-header baseline control are unchanged. A new test records that
the guard refuses today's src where it differs from the pin.
Missing history skips with the reason, or fails when
CELERIS_REQUIRE_HISTORY=1 (release gating) via history_unavailable().
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Save -> load into a fresh Simulation must restore populations (bitwise),
flags, body and step counts, and both copies must stay bitwise identical
after 40 further steps. Jets remain non-checkpointable (documented).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
run_case --current-checkout records the clean checkout (HEAD, src
fingerprint, harness hashes, pinned build-capture helper) labelled
current_checkout instead of requiring the frozen a135252 tree; receipts
carry numerical_revision=None. assess --current-checkout applies the same
gates to such receipts. Frozen mode is unchanged and now rejects any
receipt with source_mode, so the two never cross-qualify. A missing a135252
object is reported as unavailable history instead of a source diff.
check_projected_pairs.py is untouched, so its pinned hash and existing
receipts keep their meaning.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
plan.json and result.json now carry HEAD, the dirty flag, the src tree
hash and the runner hash, labelled current_checkout; result.json is
refused if the checkout changed during the run. Numerics, sampling and
the frozen steady rule are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Historical-evidence pins (dfg_re100 / jet_circle_cv / refined_dfg manifests)
now compare each recorded digest with the file at the recorded commit, read
with git, instead of with today's working tree. Missing history (no git,
shallow clone, installed copy) raises HistoryUnavailable and the test is
reported as skipped with that reason, never passed. Receipt/plan logic tests
take the selected manifest directly, so their assertions run again; the
pinned modes' mismatch rejections stay as negative controls.
V1 (run_jet_circle_cv) keeps its pinned mode and gains --current-checkout:
it records HEAD, the src tree and per-file hashes labelled as current
checkout, never as the frozen qualification, and requires a clean tree.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Supersedes 71221837. Samples every t* 0.5 (interval-mean forces), trailing
t* 50..60 window (21 samples); range <= 10 % and |slope*window| <= 2 % of
each official band width. Verdict on the window mean with the distance to
the nearest band edge. Our engineering rule, frozen before any run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 869d60d78e)
Samples every 0.5 s at synchronized root steps; trailing 5 s window (11
samples); range and |slope*window| each at most 1 % of the official band
width. States the fluid-on-cylinder force sign, reports cd_positive,
defaults to the refined case and uses the review's label.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 71221837f9)
Reuses the pinned Re100 runner and helpers unchanged; sets only the
lattice viscosity for Re 20. Reports Cd, Cl and p_diff (dp * 0.04 for
U = 0.2 m/s), a steady rule frozen before any run (window ranges at most
a tenth of each official interval width) and no interval verdict when
not steady. Declares the pressure-NEQ outlet and Zou-He inlet deviations.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 9b8ff90609)
baba5e0 replaced _advance_child with _advance(index, parent_buffer) driven
from _hop; the two feed-order tests still patched the removed method.
Patch _advance and record only index 2 (L1 hop feeding L2); assertions,
expected order and the stale-feed negative control are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
(cherry picked from commit 3f47436073)
Integrate reviewed test changes from 7c1517de819741907fb945ea0ab7864734fa3856 and 340ba4e1dc6386581839c0b28c8fe8a1f4aeb70b unchanged. Document bounded 2D implementation, evidence and open gates; production source and historical manifests are unchanged.
A (per-base read-and-sum per 80-step interval), W (window + step(80) +
one read) and W1 (window + 80 x step(1)) over equal simulated duration
from the same fresh state and warm-up, 3 interleaved repeats. Bitwise
populations and per-interval total impulses (documented engineering
tolerance) are compared outside timing; health checks untimed. Records
candidate source identity (revision, dirty flag, file hashes) and does not
use or modify frozen refined-DFG manifests/results.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Separate GPU class for a shifted MRT, all-blocked, stationary-body case
without Jets: window totals vs summed per-base reads (substep-scaled
tolerance), bitwise populations vs a raw-only stepper, a single step(N)
stepper with bitwise-equal window rows and populations, step(0)
non-change, and reset after a window. SRT/Jet test stays separately
scoped. Test-only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Scale the window-vs-summed-reads tolerance by the actual absolute fine
substep contributions (last_raw after every fine hop), labelled a
conservative engineering tolerance rather than a proof. Compare
populations by finite FP32 bits plus NaN masks and reject infinities.
Test-only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
FixedPatchStepper(observe_window_exchange=True), which requires
observe_raw_exchange, allocates a second per-link FP64 ledger per body
level. While a window is open, the unchanged BlockAccumulateExchange
kernel also accumulates into it after each fine hop. Small API:
start_exchange_window() zeroes and opens at the current completed step;
read_exchange_window(body_id, level=) is non-destructive and reports
impulse with rates over the integer window duration. initialize/
_initialize_raw and step failure invalidate; remap builds a new instance
with no open window; Jet amplitude updates may stay in-window. Default
disabled: no allocation and no extra launch; last_substep/last_base_step
unchanged. summarize_exchange gains window='window' with duration.
Host contract tests (mocked driver) and a GPU gate against summed
per-base reads with a reordering-bound FP64 tolerance.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Simulation.get_ddf() is flat (9*ny*nx); view it as (Q, cells) before
applying the fluid mask, with a flat-DDF regression. Label A/E as
runner-like force-read workloads (values discarded, no aggregation or
validation; no macro/probe/CSV cost; not benchmark throughput) and split
status lines into CONSTRUCTED and WARM_READY. Test/doc only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Untimed physical-leaf/fluid health after warm-up and after each timed
window using the existing runner limits (finite macros, rho in [0.5,1.5],
max Mach <= 0.2); refined masks as in run_refined_dfg. Parity comparator
rejects unequal level counts and any infinity; matching NaN stays allowed
only as a storage mask. Regression tests for both. No numerical gate or
production change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Explicit stream synchronization before/after each timed loop (refined
FixedPatchStepper stream, uniform Simulation stream, in their contexts);
required --max-wall-seconds outer timeout checked between variants and in
per-step loops; untimed finite leaf/fluid health checks; force notes for
B/C/E' (no host reads; final base interval only; whole-window average with
different summation order). Parity stays A/B/C vs A and E/E' vs E only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Wall-time A/B/C/E/E' variants over equal simulated duration, reusing
run_refined_dfg setup functions. Fresh state and identical warm-up per
repeat, 3 interleaved repeats, construction/JIT, warm-up and bitwise
population/NaN-mask parity checks outside timed regions. B, C and E' are
labelled timing-only (no valid force history). No numerical or
production change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Remove MRT from the FixedPatch2D rejection matrix (TRT and other
rejections retained); convert host MRT tests to stdlib unittest; add GPU
gates for unshifted dense-vs-blocked MRT two-level roots (exact bits) and
a shifted all-blocked three-level MRT smoke checking per-level omega.
No numerical source change.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Additive compile-time COLLISION_MODEL == 2 branches in BlockStep,
BlockChannelStep and patch_collide repeat the uniform collide_dispatch
MRT sequence (zero forcing, omega clamp, collide_mrt_no_force). SRT
statements, including the shifted retained-delta/direct-momentum path,
are unchanged under the alternate branch. Guards accept SRT or MRT only;
TRT, LES, non-FP32, non-double-buffer and shifted dense/mixed refinement
stay rejected. Transfers, subcycling, BFL, raw ledgers, channel closures,
shifted-Jet SRT guard and PeriodicBlockStepper are untouched. Existing
MRT operator, shifted representation and fixed kinetic rates are reused.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TRON