Robert Allan JamesandClaude Sonnet 5
8d8f3aaae2
Punch list §25 item 0.5 complete.
irq_spx now branches (one instruction, well inside the 128-byte vector slot)
to irq_spx_trampoline, a 672-byte-frame save/restore sequence that calls a
C handler and returns via ERET. The other fifteen vectors are untouched,
still routing to the existing fatal handler.
EL selection (B3) happens once, in aarch64_install_vectors(), not per
interrupt: aarch64_current_el() (item 0.4) picks VBAR_EL1 or VBAR_EL2, and
the same answer is cached in a byte flag (el2_mode_flag) that the trampoline
reads to choose ELR_EL1/SPSR_EL1 vs ELR_EL2/SPSR_EL2 -- the two forms are
genuinely different MRS/MSR encodings, not runtime-selectable operands, so
this is the cheapest correct design: decide once at install time, branch
twice (save, restore) per interrupt afterward. VBAR_EL1 was previously
written unconditionally; this closes that half of item 0.4's known gap.
EL2 is coded from the architecture reference and cannot be boot-tested in
this environment (QEMU's aarch64 virt/EDK2 combination here yields EL1) --
reported as unverified rather than asserted as tested.
FP/SIMD save is not optional (B2, carried from item 0.4's finding that the
build has no -mgeneral-regs-only): the AAPCS64 caller-saved set -- v0-v7,
v16-v31, full 128 bits each -- plus FPSR/FPCR is saved and restored around
the C handler call. v8-v15 are callee-saved by the ABI and deliberately
excluded: the handler, being ordinary compiled C, preserves those itself.
aarch64_irq_handler() (interrupts.c) is deliberately empty. Distinguishing
which interrupt fired needs the GIC's IAR, which does not exist until item
0.6; nothing unmasks or routes any source to this vector yet, so the
function is not reachable during a normal boot. Per the item's own text,
no attempt was made to manufacture an interrupt to exercise this path early
-- 0.6 (GIC) and 0.7 (timer) are what prove it took and returned one.
Verified: every hand-computed frame offset (0, 16, 32 ... 640, frame size
0x2a0=672) checked against the actual disassembly of the built kernel, not
just visually reviewed -- save and restore sequences mirror exactly, and
aarch64_install_vectors' branch on the detected EL, the flag write, and the
trampoline's read of the same flag address all confirmed consistent. Boots
clean on real QEMU output, no regression: dict_hash 0x3d4e1daf289da94f
unchanged from the item 0.1-0.4 baseline, and the item 0.4 EL banner
("AArch64: running at EL1") still prints correctly ahead of "IDT installed.".
Only aarch64-scoped files touched (isr.S, interrupts.c) -- no shared loader
or header changed, so amd64 and riscv64 are provably unaffected; not
rebuilt for this item.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
src/
Hosted StarForth VM implementation (compiled by the root Makefile).
Bare-metal kernel sources live in src/starkernel/; FORTH word
implementations in src/word_source/; the test harness in
src/test_runner/; platform shims in src/platform/.
Entry point / interpreter core
main.c— CLI entry point, VM init, DoE mode dispatch.vm.c— interpreter loop, stacks, dictionary state (the central runtime file).vm_api.c— external VM API implementation.vm_bootstrap.c— VM bootstrap initialization.vm_debug.c— debugging utilities.vm_time.c— time-related VM operations.vm_internal.h— internal-only declarations shared across thevm_*.cfiles, not part of the publicinclude/vm_api.hsurface.repl.c— REPL read-eval-print loop.cli.c— CLI argument parsing.io.c— I/O operations.log.c— logging infrastructure.
Memory / dictionary / blocks
memory_management.c— dictionary allocator.dictionary_management.c— dictionary allocation and search.dictionary_heat_optimization.c— Loop #1 execution-heat tracking.word_registry.c— word registration system.block_subsystem.c— logical→physical block mapper.blkio_file.c,blkio_ram.c,blkio_factory.c— block I/O backends (file-backed, RAM-backed) and the factory that selects between them.stack_management.c— stack operations.
Physics-driven adaptive runtime (7 feedback loops)
physics_runtime.c— main physics coordinator.physics_hotwords_cache.c— Loop #1 hot-words caching.physics_metadata.c— per-word metadata tracking.physics_pipelining_metrics.c— Loop #4 word-transition prediction.physics_execution_hooks.c— execution instrumentation.rolling_window_of_truth.c— Loop #2 circular execution-history buffer.inference_engine.c— Loops #5/#6, statistical inference (window-width, decay-slope).ssm_jacquard.c— L8 Jacquard steady-state mode selector; consumescompudynamics.cfor tuning-word/config lookups.compudynamics.c— generic compudynamics module (cd_tuning_word(),cd_tuning_vm()); the score/UCB/reward/weight constants for the L8 adaptive table live here, not inssm_jacquard.c.heartbeat_export.c— heartbeat metrics export (CSV export function itself not yet implemented — seedocs/working/architecture/heartbeat_csv_export.md).
Math / measurement
math_portable.c— portable math functions.profiler.c— performance profiling.doe_metrics.c— Design of Experiments metrics (2^7 factorial).
Any .bak file alongside a .c file here (doe_metrics.c.bak,
inference_engine.c.bak, vm.c.bak) is a pre-edit backup left by a past
maintenance script (see scripts/remove_loop_conditionals.sh), not a
build input.