Enabling the epoch serving-mode JIT globally regressed continuation-based guest
interpreters (the epoch mode is the shared command channel every loop's
conformance runner uses). Two-part fix:
1. SAFE DEFAULT GATE. register_jit_hook in the persistent server branch is now
opt-in via SX_SERVING_JIT=1 (default OFF). Default behaviour is unchanged
(no JIT in epoch serving) → zero regression for sibling loops. The
content/Smalltalk page server opts in.
2. GENERAL FIXES + per-guest interpret-only declarations:
- callable? (sx_server/run_tests/integration_tests/mcp_tree) now accepts
VmClosure. A JIT-compiled higher-order function returns its inner closure
as a VmClosure; callable? previously rejected it, so scheme-apply's
(callable? proc) guard failed with "not a procedure: <vm:anon>".
- jit-exclude! gains a trailing-"*" namespace-prefix form
(Sx_types.jit_excluded_prefixes), the robust way to mark a whole guest
interpreter interpret-only (a name-list misses functions in extra files —
it left erlang's vm/dispatcher JIT'd and 13 tests short).
- Per-guest exclusions in each guest's runtime.sx:
scheme "scheme-*" "scm-*" erlang "er-*" "erlang-*"
prolog "pl-*" common-lisp "cl-*" "clos-*"
js "js-*" haskell "hk-*"
Verified under opt-in JIT (== CEK, no hang): smalltalk 847/847, scheme/flow
166/166, erlang 530/530, prolog 590/590, apl 152/152, js 147/148. Residual
(documented, protected by the default gate): common-lisp 6 fails in advanced
suites (parser-recovery/debugger/CLOS/MOP). lua (0/16) and tcl (3/4) fail
identically on CEK — pre-existing, not JIT. run_tests --jit/no-jit unchanged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- pl-eval-arith: add floor, ceiling, truncate, round, sqrt, sign, pow, integer,
float, float_integer_part, float_fractional_part, **, ^ operators
- pl-collect-vars: helper that extracts unbound variables from a term (left-to-right,
deduplicated by var id)
- term_variables/2: dispatches via pl-collect-vars, unifies second arg with var list
- pl-predsort-insert!: inserts one element into a sorted list using a 3-arg comparator
predicate; deduplicates elements where comparator returns '='
- pl-predsort-build!: builds sorted list via fold over pl-predsort-insert!
- predsort/3: full ISO predsort — sorts and deduplicates a list using a caller-supplied
predicate
- lib/prolog/tests/advanced.sx: 21 tests (12 arith, 5 term_variables, 4 predsort)
- conformance.sh: add advanced suite
- scoreboard: 517/517 (was 496/496)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds two new builtins to lib/prolog/runtime.sx:
- sub_atom/5: non-deterministic substring enumeration. Iterates all
(start, length) pairs over the atom string, tries to unify Before,
Length, After, SubAtom for each candidate. Uses CPS loop helpers
pl-substring, pl-sub-atom-try-one!, pl-sub-atom-loop!. Fixed trail
undo semantics: only undo on backtrack (k returns false), not on success.
- aggregate_all/3: collects all solutions via pl-collect-solutions then
reduces. Templates: count, bag(T), sum(E), max(E), min(E), set(T).
max/min fail on empty; count/bag/sum/set always succeed.
New test suite lib/prolog/tests/string_agg.sx: 25 tests, all passing.
Total conformance: 496/496.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds 6 new built-in predicates to the Prolog runtime and 24 tests covering
term<->atom conversion (bidirectional), output capture, format directives (~w/~a/~d/~n/~~).
456/456 tests passing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds 5 new built-in predicates to the Prolog runtime with 15 tests.
390 → 405 tests across 20 suites (all passing).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds pl-apply-goal helper for safe call/N goal construction (atom or compound),
five solver helpers (pl-solve-forall!, pl-solve-maplist2!, pl-solve-maplist3!,
pl-solve-include!, pl-solve-exclude!), five cond clauses in pl-solve!, and a
new test suite (15/15 passing). Total conformance: 390/390.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
