`remove_dead_loads` now scans past opcodes that touch neither A nor
the flags an LDA sets, so a redundant LDA gets caught by its
successor's overwrite even when an index load or counter bump sits
between them. The extension covers LDX/LDY/INX/INY/DEX/DEY and the
flag ops (CLC/SEC/CLI/SEI/CLD/SED/CLV) alongside the INC/DEC/STX/STY
opcodes the pass already stepped past.
The highest-leverage case is the shape every single-tile `draw`
emits. After copy propagation and dead-store elimination do their
work, the stream reads:
LDA #<y> ; stray producer, value never consumed
LDY oam_cursor
LDA #<y> ; real load before STA
STA $0200,Y
The first LDA was surviving because the pass bailed on the LDY.
With the step-past, it drops. One LDA gone per draw, 2 bytes each.
Measured LDA-count reduction on committed examples:
platformer 242 → 221 (-21, -8.7 %)
war 785 → 754 (-31, -4.0 %)
pong 843 → 827 (-16, -1.9 %)
**Audio goldens.** The cycle savings shift the main-loop/NMI boundary
in audio-emitting programs, which re-times which frame each SFX
trigger lands in. Six audio hashes re-baseline as a result:
audio_demo, friendly_assets, noise_triangle_sfx, platformer, pong,
war. All 50 PNG goldens, the platformer/war/pong demo gifs, and
every non-audio program stay byte-identical. The re-baselined
output is still sample-accurate; what changed is the first-SFX
offset within the captured 132 084-sample window. This is the
audio-shift tradeoff documented in future-work.
Two new peephole unit tests lock in the behaviour:
- `dead_load_elim_steps_past_ldx_ldy` — the DrawSprite shape folds.
- `dead_load_elim_preserves_lda_when_used_by_shift` — a subsequent
ASL on A keeps the LDA alive across an intervening LDY.
Also updates future-work.md to reflect the shipped change and the
remaining register-allocator wins worth chasing next.
Closes §H. 2×2 metatiles and a parallel collision map are now a
first-class construct. `metatileset Name { metatiles: [{ id, tiles,
collide }, ...] }` declares a library of 2×2 tile bundles. `room Name
{ metatileset: M, layout: [...] }` lays them out on a 16×15 grid. The
compiler expands each room at compile time into:
- a 960-byte nametable (`__room_tiles_<name>`)
- a 64-byte attribute table (`__room_attrs_<name>`)
- a 240-byte collision bitmap (`__room_col_<name>`)
`paint_room Name` reuses the vblank-safe `load_background` update
machinery for the nametable blit and installs the collision bitmap
pointer into `ZP_ROOM_COL_LO`/`ZP_ROOM_COL_HI` (ZP $18/$19).
`collides_at(x, y)` JSRs into a small runtime helper that reads
`(room_col),Y` with `Y = (y & 0xF0) | (x >> 4)` and returns 0/1.
The helper links in only when the `__collides_at_used` marker is
emitted, so programs that declare a room but never query it pay
zero bytes for the subroutine.
`parse_byte_array` grows a `[value; count]` shortcut — 240-entry
`layout` arrays are unwieldy to spell out a byte at a time.
See `examples/metatiles_demo.ne` for the end-to-end flow: a probe
sprite bounces off walls via `collides_at` and lands on the left
side of the playfield at frame 180 — direct evidence that the
collision query works.
Also defers the register-allocator work from §"Code quality /
tooling" and documents the audio-goldens constraint in future-work
so the next agent sees it.
Closes the §A follow-up gap: ordering compares (`<`, `<=`, `>`, `>=`)
on signed integer types now use the canonical 6502 `CMP / SBC / BVC /
EOR #$80` overflow-correction idiom so the N flag reflects the true
sign of the difference, instead of the previous BCC/BCS-based path
that always treated `$FFxx` as greater than `$00yy`.
The same change also fixes narrow-to-wide widening: assigning a
runtime `i8` expression to an `i16` variable now sign-extends the
high byte via a new `IrOp::SignExtend` op instead of zero-extending
it, so `var w: i16 = some_i8_neg` round-trips negative values.
The lowerer tracks signedness on each IR temp (analogous to the
existing `wide_hi` map) and threads it onto the new `Signedness`
field of `CmpLt`/`CmpGt`/`CmpLtEq`/`CmpGtEq` and their 16-bit
variants. The optimizer's constant-folder uses the same flag to
fold compares correctly under either signedness. Casts to `u8`/`u16`
strip the signed flag so an explicit `as` opt-out stays unsigned.
`examples/signed_compare.ne` exercises both bit widths through the
emulator harness — the four pip sprites at the top of the screen
show three lit (signed-correct) and one dark (would only light if
the compare regressed to unsigned semantics).
Each $2006 write inside __vram_buf_drain updates the PPU's `t`
(scroll) register, so leaving it pointing at the last buffer
entry's address shifted the next frame's rendering up/right by
however many cells we wrote past $2000. Reset by writing $00 to
$2006 twice (clears `t` and resets the write-toggle to high)
followed by $00 to $2005 twice (zero X/Y scroll). The HUD demo
golden flips from "smileys offset by ~16px" to the intended
red bar with white hearts and a yellow score digit.
https://claude.ai/code/session_01F7dHsgh7UX7SAK3wZ7JiKc
Closes the highest-priority remaining catalogue item (§G). User
code queues PPU writes during `on frame` via three new intrinsics;
the NMI drains the 256-byte ring at `$0400-$04FF` to `$2007`
during vblank. Programs that never touch the buffer pay zero
bytes and zero cycles for the feature — verified by the existing
46 ROMs all matching their goldens with no drift.
Also fixes the failing CI Format check from 7b4570e by running
cargo fmt across the working tree.
**Runtime:**
- New `runtime::gen_vram_buf_drain` emits the drain routine
(`__vram_buf_drain`). Walks entries `[len][addr_hi][addr_lo]
[byte_0]...[byte_(len-1)]` and stops at `len == 0`. Uses
`LDA $0400,X` indexed-absolute so no ZP scratch is needed.
Drain costs ~12 setup cycles + 8 cycles per data byte; the
256-byte buffer can hold ~50 single-tile writes that drain
in roughly 1000 cycles, well inside the ~2273-cycle vblank.
- `NmiOptions` gains `has_vram_buf`. The NMI JSRs the drain
after the existing palette/background handshake (compiler-
queued PPU writes win priority for vblank cycles).
**IR + codegen:**
- Three new ops `IrOp::NtSet`, `IrOp::NtAttr`, `IrOp::NtFillH`.
- The codegen helpers compute the PPU address inline:
`$2000 + y*32 + x` for nametable, `$23C0 + (y/4)*8 + (x/4)`
for attribute. Each append lays down a fresh `0` sentinel so
the NMI sees a well-formed buffer regardless of whether more
entries get appended later in the frame.
- `__vram_buf_used` marker drops on first use; gates the
runtime splice + NMI JSR.
**Analyzer:**
- AST-walking helper `program_uses_vram_buf` detects intrinsic
use at analyze-init time so the user-RAM bump pointer can
start at `$0500` (past the buffer) rather than the legacy
`$0300`. Programs that don't use the buffer keep the legacy
start.
- Three intrinsic names registered in `is_intrinsic` /
`is_void_intrinsic` with arity checks.
**Tests + example:**
- `examples/vram_buffer_demo.ne` exercises all three intrinsics
on a backgrounded program — three single-tile score writes,
a 16-tile horizontal fill, and an attribute write that flips
the top-left metatile group's palette to red. Committed
golden + audio hash.
- Four new integration tests: byte-level JSR-to-drain
assertion, drain-omitted-when-unused, RAM-bump assertion for
programs that DO use the buffer, and arity enforcement for
`nt_set`.
**CI fix:**
- `cargo fmt` ran across the tree. Picks up a one-line fmt
diff in `tests/integration_test.rs` that the prior commit
shipped without running fmt, causing the Format CI job to
fail on `7b4570e`.
All 758 tests pass. Clippy clean. 47/47 emulator goldens match.
Another batch from the cc65/nesdoug catalogue. All gated on
parser-level opt-in or default-false attributes so existing
programs produce byte-identical ROMs (no committed .nes file
changed).
**§A — `i16` signed 16-bit type:**
- New `KwI16` lexer token, `NesType::I16` AST variant, parser
case in `parse_type`. Type-size and integer-type tables
treat `i16` like `u16` (2 bytes, integer).
- IR lowering accepts `i16` everywhere it accepts `u16` for
wide-load / wide-store / widen-narrow paths.
- New constant fold for `UnaryOp::Negate(IntLiteral(v))` that
emits the wide two's-complement form. Without it, `var vy:
i16 = -10` would zero-extend to `$00F6` (= 246) instead of
sign-extending to `$FFF6` (= -10). Negative literals now
store the right bytes.
- Comparisons reuse the existing unsigned 16-bit compare ops
(matching the existing `i8` behaviour). Documented in the
`NesType::I16` doc comment and in `future-work.md` §A.
- Example `examples/i16_demo.ne` with committed golden.
- Tests cover the literal-fold sign-extension and end-to-end
compile of the example.
**§S — SRAM / battery-backed saves:**
- New `save { var ... }` top-level block. Lexer + parser opt
into a dedicated `KwSave` token. Analyzer allocates save
vars from a separate `next_sram_addr` bump pointer starting
at `$6000`, capped at `$8000` (8 KB cartridge SRAM window).
- Linker reads `analysis.has_battery_saves` and flips iNES
byte-6 bit-1 via the new `RomBuilder::set_battery` /
`Linker::with_battery` chain.
- New `W0111` warning for save-var initializers — SRAM is
preserved across power cycles, so an init expression would
either silently never run or clobber persisted data on
every boot. The warning teaches the user about the
magic-byte sentinel pattern.
- Struct fields in save blocks are explicitly rejected for now
(the field-flattening path uses the main-RAM allocator).
- Example `examples/sram_demo.ne` with committed golden, plus
4 integration tests.
**§D (partial) — inline-asm `.label:` syntax:**
- Codegen-side mangler rewrites `.IDENT` → `__ilab_<N>_IDENT`
per inline-asm block, where `<N>` is the call site's
monotonic suffix. Two `asm { .loop: ... }` blocks in the
same function now coexist without colliding in the linker's
label table.
- Bounds checks on `.` placement: `$2002` and `name.field`
are unaffected; only `.IDENT` in label / branch context
triggers the rewrite. Two integration tests pin the
uniqueness and dollar-vs-dot disambiguation.
**§X follow-up — Mesen trace-log docs:**
- New "Debugger-assisted workflows" section in
`docs/nes-reference.md` walking through the Mesen / FCEUX
log workflows alongside the new `debug_port:` attribute.
**Misc:**
- `future-work.md` updated to mark the shipped items out of
the catalogue and reshuffle the priority ranking. Remaining
niche follow-ups (signedness on Cmp16, struct save fields,
inline-asm format specifiers) documented inline so future
passes know the design.
All 757 tests pass. Clippy clean. 46/46 emulator goldens match.
Another batch from the cc65/nesdoug gap catalogue. All six items
gated on marker labels (or default-false attributes) so existing
programs produce byte-identical ROMs — every pre-existing .nes
file round-trips unchanged.
**Language / runtime additions:**
- `mapper: GNROM` (iNES 66). Combines AxROM's 32 KB PRG pages with
CNROM's 8 KB CHR banks in a single `$8000` register. Linker
pads single-page ROMs to 32 KB to match mapper-66 expectations.
- `game { debug_port: fceux | mesen | 0xXXXX }`. `debug.log`,
`debug.assert`, and the `__debug_halt` sentinel now target a
user-selected address. `fceux` (default, $4800) and `mesen`
($4018) are named aliases; custom hex addresses are accepted
for unusual debuggers.
- `game { sprite_flicker: true }`. IR lowerer injects an
`IrOp::CycleSprites` at the top of every `on frame` handler,
which flips on the rotating-OAM NMI variant with no per-site
boilerplate. Default false so existing ROMs keep their layout.
- `fade_out(step_frames)` / `fade_in(step_frames)` builtins.
Blocking helpers that walk brightness 4 → 0 or 0 → 4 with
`step_frames` frames between each step. Runtime splices
`__fade_out`, `__fade_in`, and a callable `__wait_frame_rt`
helper when the builtin is used. Zero-guard on step_frames
prevents a pathological 256-frame spin when the caller
accidentally passes 0.
- `sprite_0_split(scroll_x, scroll_y)` intrinsic. Emits a
two-phase busy-wait on `$2002` bit 6 (wait-for-clear,
wait-for-set) then writes the new scroll values to `$2005`.
Works on any mapper — unlike `on_scanline(N)` which requires
MMC3. Enables HUD-over-playfield scrolling on NROM/UxROM/MMC1.
**Docs:**
- New paragraph in the language guide explaining the no-recursion
design choice and the explicit-stack workaround pattern.
- `future-work.md` updated to mark the shipped items out of the
catalogue; remaining items reshuffled in the priority ranking.
- README + examples/README updated with the new mapper and
builtins.
**Tests:**
- 12 new integration tests covering: GNROM header emission,
debug-port targeting (fceux/mesen/custom), unknown-alias
rejection, sprite_flicker on/off/bad-value, fade_out JSR + marker
coupling, fade omitted-when-unused, fade-in-expression rejected,
sprite_0_split byte-level busy-wait verification, sprite_0_split
arity enforcement, sprite_0_split omitted-when-unused, and an
extended void-intrinsic-in-expression-position test covering the
three new void builtins.
- `nes2_mapper_high_nibble_in_byte_8_is_zero_for_small_mappers`
extended to include GNROM.
- Four new examples with committed .nes ROMs + pixel/audio
goldens: `gnrom_simple`, `auto_sprite_flicker`, `fade_demo`,
`sprite_0_split_demo`.
All 752 tests pass. Clippy clean. 44/44 emulator goldens match.
Follow-up cleanup on the cc65 parity batch. Addresses issues found
during a post-commit code review.
**Correctness fixes:**
- `rand8()` / `rand16()` at statement position (result discarded)
were being eliminated by DCE because `op_dest` returned
`Some(dest)` for Rand8/Rand16 even though the ops have a visible
side effect — advancing the PRNG state. Now `op_dest` returns
`None` for both, keeping the JSR regardless of liveness. New
regression test `rand8_statement_survives_dce`.
- Void-only intrinsics (`poke`, `seed_rand`, `set_palette_brightness`)
used in expression position (e.g. `var x = seed_rand(42)`) were
panicking the linker with an unresolved `__ir_fn_X` label. The
analyzer now emits E0203 with a clear message; new
`void_intrinsic_in_expression_position_errors` test covers all
three names.
- Statement-position `rand8()` / `rand16()` weren't lowered at all
(they fell through to the default Call path). Now both lower to
their IR op with a fresh temp that nothing reads; the JSR still
runs so the PRNG state advances.
- `--fceux-labels foo.nes` was producing `foo.0.nl` because
`PathBuf::with_extension` replaces instead of appends. Rewritten
to literally append `.<bank>.nl` / `.ram.nl` to the OsString, so
users get the FCEUX-expected `foo.nes.<bank>.nl` naming.
- Linker now asserts CNROM / AxROM don't accept user-declared
switchable PRG banks — their page sizes don't fit the 16 KB per
bank model, and silently producing a mis-sized ROM is worse than
a loud panic.
**PRNG cleanup:**
- Removed the stream-of-consciousness comment block in `gen_prng`
that described three abandoned algorithms before landing on the
actual Galois LFSR.
- Simplified `__rand16` to a single JSR + LDX instead of two
JSRs + TAY/TYA round-trip — a single shift already produces 16
fresh bits, the doubled call just burned ~40 cycles. The golden
PNG for `prng_demo` was regenerated to reflect the new sequence.
- Rewrote the `gen_prng` doc comment to accurately describe the
algorithm as a Galois LFSR (it was mislabelled as xorshift).
- Rewrote the `gen_palette_brightness` doc comment with a proper
table of level→mask mappings — the prior prose description
didn't match the actual table values.
**Tests:**
- Three new unit tests in `linker::debug_symbols` covering the
FCEUX `.nl` renderer: user-facing labels only, empty output when
no user labels exist, and deterministic sorting in `.ram.nl`.
- Extended `nes2_mapper_high_nibble_in_byte_8_is_zero_for_small_mappers`
to cover AxROM + CNROM.
- Renumbered priority list in future-work.md after removing the
shipped sections (J, K, N, parts of V and Y).
All 737 tests + 40/40 emulator goldens still green.
Closes seven of the cc65/nesdoug parity gaps catalogued in
docs/future-work.md in a single pass. All of the new features are
gated on marker labels so programs that don't use them produce
byte-identical ROM output (every pre-existing committed .nes file
round-trips unchanged).
Language / runtime additions:
- `rand8()` / `rand16()` / `seed_rand(u16)` intrinsics backed by a
16-bit Galois LFSR (~30 bytes of runtime, ~40 cycles per draw).
Reset path seeds state to 0xACE1 so the first draw is useful even
without explicit seeding.
- `p1.button.a.pressed` / `.released` edge-triggered input via a
new ReadInputEdge IR op plus an NMI-side prev-frame snapshot into
$07E6/$07E7, gated on the `__edge_input_used` marker.
- `set_palette_brightness(level)` builtin mapping levels 0..8 to
PPU mask emphasis bytes (`$2001`) for neslib-style screen fades.
- `mapper: AxROM` (iNES 7) with automatic 32 KB PRG padding so
emulators that enforce mapper-7's 32 KB page size boot cleanly.
- `mapper: CNROM` (iNES 3) with a reset-time CHR bank 0 select.
- `--fceux-labels <prefix>` CLI flag emitting per-bank `.nl` label
files and a `.ram.nl` file for FCEUX's debugger.
Tests + examples:
- Five new example programs with committed .nes ROMs and
pixel+audio goldens: prng_demo, edge_input_demo,
palette_brightness_demo, axrom_simple, cnrom_simple.
- Seven integration tests covering JSR emission, the
omitted-when-unused invariant, the NMI prev-input snapshot, the
correct mapper numbers for AxROM/CNROM, and negative tests for
unknown button names and bad rand8 arity.
- `is_intrinsic()` now runs in expression-position Call paths too,
so `var x = rand8(1, 2)` errors at compile time instead of
silently dropping the extra arguments.
Follow-up to the silent-drop audit. The old ABI passed every
parameter through four fixed zero-page transport slots `$04-$07`,
imposing a hard 4-param cap (E0506) that didn't compose with
structs/arrays/u16s and fell back to "pack args into a global"
workarounds whenever a function needed five things. The transport
scheme also cost every non-leaf call a 4-LDA/STA spill prologue
(~28 cycles, 16 bytes) to copy args out of ZP before the next
nested `JSR` could clobber them.
Replace it with a hybrid convention keyed on leaf-ness:
- **Leaf callees** (no nested `JSR` in body, ≤4 params):
unchanged. Caller stages args into `$04-$07`; body reads those
slots directly for its entire lifetime. No prologue copy.
Fastest path, 3-cycle ZP stores + 3-cycle ZP loads, preserves
the SHA-256 leaf-primitive optimisation that motivated the
original fast path.
- **Non-leaf callees** (body contains a nested `JSR`, OR ≥5
params): direct-write. Caller stages each argument straight
into the callee's analyzer-allocated parameter RAM slot,
bypassing the transport slots entirely. No prologue copy on
the callee side. Saves ~24 cycles and ~16 bytes per call vs
the old transport-then-spill path, and — crucially — scales
past 4 params because the per-param slots live wherever the
analyzer put them rather than in a fixed ZP window.
The analyzer's ceiling moves from 4 to 8. Functions with 5–8
params are silently promoted to the non-leaf convention (even if
their body has no nested `JSR`), which pays the direct-write cost
rather than the prologue-copy cost — still cheaper than the old
ABI. Declarations with 9+ params still emit E0506.
### Implementation
- `function_is_leaf` now also requires `param_count <= 4`.
- `IrCodeGen::new` populates `non_leaf_param_addrs: HashMap<String,
Vec<u16>>` — for every non-leaf function, the ordered list of
addresses its parameters occupy. Callers use this to route each
arg directly to the right slot.
- `IrOp::Call` branches on presence in the map: non-leaf → direct-
write, leaf (or absent — 0-arg case) → ZP transport.
- `gen_function` no longer emits a prologue. Leaves didn't have
one; non-leaves had a 4-LDA/STA copy that is now unnecessary
because args arrive pre-written to the slot.
- The previous `leaf_functions: HashSet<String>` field is
removed; leaf-ness is now inferred from absence-in-
`non_leaf_param_addrs` at the call site.
### Tests and regressions
- `eight_param_non_leaf_function_stages_every_arg_at_its_allocated_slot`
compiles an 8-param function, scans PRG for a distinct
`LDA #\$NN / STA <addr>` per arg (immediates `0x11..0x88`), and
asserts that STAs to the `$04-$07` range are strictly fewer
than 8 — proof the old transport path is gone for this call.
- `non_leaf_call_direct_writes_args_to_callee_param_slots`
replaces the old `gen_function_prologue_spills_params_to_local_ram`
test with a dual assertion: (a) no `LDA \$04` prologue at the
callee entry, and (b) the caller-side STA lands at the
analyzer-allocated param slot, not at `\$04-\$07`.
- `analyze_rejects_function_with_more_than_4_params` renamed and
rewritten for the new 8-param cap.
- `feature_canary.ne` gains a 6-param `sum6` call (1+2+3+4+5+6 =
21) as check 8. The canary stays green (all eight checks
pass), so the committed golden is unchanged.
### Blast radius
- Six example ROMs change bytes (arrays_and_functions, function_chain,
mmc1_banked, pong, sha256, war) because their non-leaf call sites
pick up the shorter staging sequence.
- Pong and war audio hashes refresh (pure layout-timing shift; no
behavioural change in the 180-frame no-input window). docs/pong.gif
and docs/war.gif stay byte-identical.
- `examples/function_chain.ne`'s header comment updated to
document the leaf vs non-leaf split it exercises.
- `docs/language-guide.md` parameter-count section and E0506 entry
updated to reflect the new rule.
All 720 Rust tests pass; all 35 emulator goldens pass.
https://claude.ai/code/session_01AoQ678uVeqpyayvWHpfDhC
Phase 3/4 of the post-PR-#31 audit.
### Call args > 4 is now an assert
`IrOp::Call` silently `.take(4)`-d the arg list with a comment
claiming the analyzer's E0506 check made the extras unreachable.
Replace with an explicit `assert!(args.len() <= 4, ...)` so if
the analyzer ever regresses, the codegen crashes loudly instead
of miscompiling the call. Iterate over all args (not just the
first 4) since the assert guarantees correctness.
### CLAUDE.md: new-feature PR checklist
Document the lesson the audit taught: every new language-feature
PR must include (1) an example exercising it, (2) a runtime
*behaviour* assertion (not just a "ROM validates" shape check),
(3) a negative test for invalid use. Call out the specific
address-map lookup pattern (`if let Some(&addr) = map.get(..)`
with no else) that shipped the state-local bug, and recommend
the `IrCodeGen::var_addr` / explicit `.unwrap_or_else(|| panic!)`
idiom instead.
Chose not to add a regex-based CI tripwire for "silently" /
"for now" comments because the false-positive rate against
legitimate design decisions ("silently truncate to 8 bits per
the cast spec", etc.) would train contributors to ignore it.
The durable checklist in CLAUDE.md is what next agents need.
https://claude.ai/code/session_01AoQ678uVeqpyayvWHpfDhC
Phase 2 of the post-PR-#31 audit. The codebase had four documented
"silently skip" paths that parsed user intent but produced no code.
Each one was the same shape as the state-local bug: the analyzer
accepted the program, the IR lowered the construct, but somewhere
downstream the emitted code was dropped on the floor — and a pixel
golden that captured the broken behaviour locked it in as correct.
Fix each per the plan, either by implementing the feature or
rejecting the program at the analyzer.
### on_exit handlers now actually run
`IrOp::Transition` used to comment "on_exit of the current state
isn't called here because we don't know from an IR op alone which
state we're leaving." The codegen emitted the exit handler's body
as an IR function but never JSR'd it. Three example programs
(pong, war, state_machine) relied on `stop_music` or mode-flag
translation inside `on exit` that had been silently never running.
Emit a small CMP-chain against `ZP_CURRENT_STATE` before each
transition: for every state that declares an on_exit, compare the
current index, branch past on miss, JSR the exit handler on match,
then JMP to the shared done-label so only the leaving state's
handler fires. The chain is inlined at each transition site
(bounded by the number of states declaring on_exit) rather than
factored into a single trampoline — simpler to reason about, and
transitions are rare enough that the extra bytes don't matter.
Pong / war / state_machine ROMs change because the dispatch code
is now emitted. Video goldens stay byte-identical (no transitions
happen within the 180-frame harness window under no-input). Pong
and war audio hashes shifted from pure code-layout timing and are
regenerated. `docs/pong.gif` and `docs/war.gif` are byte-identical.
### State-local array initializers now refuse to compile (E0601)
`src/ir/lowering.rs:887` had the comment "Array initializers for
state-locals aren't supported yet... Programs that try this should
get a diagnostic from the analyzer; for now, silently skip." The
analyzer never actually emitted that diagnostic. Verified by
compiling `state Main { var buf: u8[4] = [10,20,30,40] ... }`:
the program built a valid ROM with no trace of 10/20/30/40 in PRG.
Add E0601 to the analyzer's state-local pass. The IR lowerer's
defensive `continue` stays in place as a belt-and-braces guard.
### `on scanline` without MMC3 is now E0603
Previously E0203 ("invalid operation for type") which is a
miscategorisation — the feature is unsupported on the current
mapper, not a type error. Dedicated E0603 makes the future-work
shape explicit.
### `slow` variables now actually live outside zero page
`Placement::Slow` was parsed into the AST but `allocate_ram`
ignored it, so `slow var cold: u8` still landed in ZP like any
other u8. Wire `var.placement` through `allocate_ram_with_placement`
and skip the ZP branch when `Slow` is set. `Fast` remains
advisory (the existing default already prefers ZP for u8 vars),
validated by W0107.
### Other address-map silent drops hardened
Alongside the var_addrs hardening from phase 1, three `state_indices`
lookup sites that did `.copied().unwrap_or(0)` or silent `if let`
are now explicit panics: scanline IRQ dispatch, MMC3 reload, and
`IrOp::Transition`. A miss in any of them is a compiler bug, not
valid input — the analyzer catches unknown state names upstream.
### Regression guards
Four new tests would have failed against the old silently-dropping
code paths:
- `analyze_state_local_array_initializer_rejected` — expects E0601.
- `analyze_on_exit_declaration_accepted` — expects no errors.
- `analyze_slow_var_forced_out_of_zero_page` — expects alloc
address >= $0100.
- `transition_dispatches_leaving_states_on_exit_handler` — counts
distinct JSR targets in the PRG before/after adding `on exit` to
a state; the exit-bearing build must have more.
All 720 tests pass. All 34 emulator goldens pass after the pong/war
audio hash refresh.
https://claude.ai/code/session_01AoQ678uVeqpyayvWHpfDhC
Phase 1 of the post-PR-#31 audit. The PR #31 state-local bug had a
specific shape: analyzer allocated a slot, codegen looked it up by
VarId, silently emitted nothing on miss. Six sites in gen_op plus
the global-initializer loop and the parameter-shuffle prologue all
used the same `if let Some(&addr) = self.var_addrs.get(var) { ... }`
pattern with no else branch. Any future allocation-map desync would
slip through the same crack.
Replace every site with a new `IrCodeGen::var_addr(VarId) -> u16`
helper that panics with an explicit "compiler bug" message on miss.
An IR op referencing an unmapped VarId is not valid input — it means
the analyzer and lowerer disagreed on what to allocate, and we want
that crash to surface in CI rather than be absorbed by whatever
zero-filled RAM happened to sit at the read.
Running cargo test against the hardened lookup surfaced exactly the
bug shape the plan predicted: uninitialized struct globals (e.g.
`var p: Point` with no literal initializer) never had their flattened
field VarIds (`"p.x"`, `"p.y"`) registered in var_addrs. The IR
lowerer's `get_or_create_var("p.x")` minted a VarId, the analyzer's
`flatten_struct_fields` allocated an address for it, but IrCodeGen::new
only populated var_addrs from `ir.globals`, which doesn't contain
synthesized field entries for uninitialized structs. Every `p.x = N`
silently compiled to nothing.
Fix by exposing the IR lowerer's name→VarId map on IrProgram and
joining it with the analyzer's allocations in IrCodeGen::new. Every
allocated name that the lowerer knows about now gets a var_addrs
entry. Example ROMs are byte-identical (no example relied on the
dropped writes), but the bug was reachable — any user program with
a plain `var pos: Point` declaration and field writes would have hit
it silently.
Add `uninitialized_struct_field_store_emits_sta_to_allocated_address`
as a byte-level regression guard: compile `p.x = 123` and scan PRG
for `LDA #\$7B / STA <addr>`. Fails against the old silently-dropping
codegen.
https://claude.ai/code/session_01AoQ678uVeqpyayvWHpfDhC
Before this change, state-local variables (`state Foo { var x: u8 = 0 }`)
were silently no-ops: the analyzer allocated a ZP slot for them, but
the codegen's `var_addrs` map only covered IR globals and scope-qualified
function locals — so every `LoadVar` / `StoreVar` whose `VarId` pointed
at a state-local resolved to no address and emitted nothing. Existing
examples compiled and matched their goldens because none of them observed
the dropped writes within the 180-frame harness window.
The overlay changes the analyzer's state-local pass to snapshot both the
ZP and RAM cursors after the globals have been laid out, then rewind to
that snapshot before each state's locals and track the running max.
`ZP_CURRENT_STATE` keeps exactly one state active at runtime, so every
state's locals are mutually exclusive with every other state's and can
share the same bytes. The IR lowerer now pushes each state's locals into
the IR globals table (with `init_value=None`) so the codegen resolves
their addresses the same way it does program globals, and prepends the
declared initializers to each state's `on_enter` handler (synthesizing
an empty one where needed) so a freshly-entered state re-establishes its
bytes before user code runs.
`--memory-map` now tags each allocation with its owning state
(`[@Title]`, `[@Playing]`, ...) and counts distinct bytes rather than
summed allocation sizes so overlaid slots don't double-count. The
`AnalysisResult.state_local_owners` map exposes the ownership to any
tool that wants to group allocations the same way.
Only `state_machine.ne` and `platformer.ne` declare state-level vars,
so they're the only example ROMs whose bytes change. `platformer.ne`'s
audio golden shifts slightly (the now-working `blink` counter in Title
adds a few cycles per frame before the auto-transition to Playing, which
offsets APU register writes within each frame); its video golden and
every other example ROM stay byte-for-byte identical.
Fixes#22.
https://claude.ai/code/session_015kvJu3iEFLSRJoShPBfm3X
The `seg.ooffs` field in our ca65 .dbg output was off by 16 — it was
emitting the PRG-relative fixed-bank offset when ca65's convention
(and Mesen's DbgImporter.cs:301 math:
`Address = val - seg.start + ooffs - headerSize`) expects the raw
output-file offset, *including* the iNES header. The practical
consequence: every label Mesen resolved via the .dbg was 16 bytes
short of its true PRG offset, which silently corrupted source-line
mapping for the first bytes of each function.
Fix is a one-liner — drop the `saturating_sub(16)` and feed
`linked.fixed_bank_file_offset` straight into the ooffs field. Unit
tests in debug_symbols.rs updated to assert the new values (ooffs=16
for NROM, 16+16K*N for banked).
The Mesen probe (`tests/mesen/probe.lua`) is expanded in the same
change, because the sabotage test that caught this bug is also the
cleanest demonstration the probe is working:
* checks all four entry-point labels resolve and land inside the
fixed bank's CPU window ($C000-$FFFF);
* asserts the linker's relative ordering (main_loop < Main_frame
< nmi);
* registers a startFrame callback, waits three frames, and verifies
PC is still in the fixed bank + that `emu.read(main_loop.address,
nesPrgRom)` returns 0xA5 (the LDA-zp opcode the runtime always
places as main_loop's first instruction). The 0xA5 constant is
what catches the ooffs regression — a less-specific "not 0xFF"
check coincidentally passed even with ooffs=0 because the shifted
address still landed on real code.
Verified locally by running the probe against hello_sprite's ROM
with four different `.dbg` mutations and confirming each triggers
the expected exit code.
https://claude.ai/code/session_01DfN3pKJLryr7vvNFBpcqmC
Emit a `.dbg` debug-info file in the same format `ld65` produces, so
Mesen / Mesen2 / fceuX pick it up automatically and enable source-line
stepping, labelled variable inspection, and symbol-based breakpoints
without manual address lookups. Closes#23.
The new `render_dbg` helper stitches together metadata the compiler
already surfaces (linker label table, IR codegen `__src_<N>` markers,
analyzer variable allocations) into the file/mod/seg/scope/span/line/sym
records documented at https://cc65.github.io/doc/debugfile.html. Each
source-loc marker becomes a span that stretches to the next marker
(so breakpoints cover every byte the statement compiled into) plus a
line record pointing into it; `seg.ooffs` tracks the fixed bank's
PRG-relative start so banked MMC1/UxROM/MMC3 ROMs map cleanly too.
Reuses the `.mlb` symbol-name filter so internal skip/block labels
stay out of the debugger's symbol browser. `--dbg` implies the same
`__src_` marker emission as `--source-map` but leaves release builds
byte-identical when neither flag is passed.
https://claude.ai/code/session_01DfN3pKJLryr7vvNFBpcqmC
Move the six gate-marker label emissions (__mul_used, __div_used,
__oam_used, __default_sprite_used, __p1_input_used, __p2_input_used)
out of the inline IR-op lowering paths and into a new
`emit_trailing_markers()` helper that runs once at the end of
`generate()`. The IR walk now just flips a bool per marker; the
label emit happens after every instruction has been lowered, so
the marker never lands in the middle of a peephole-sensitive
sequence.
Fixes a real peephole interaction that surfaced after rebasing on
main's `codegen: skip parameter-spill prologue for leaf functions`
+ `peephole: drop dead LDA #imm before mem-INC/DEC + JMP`
improvements: an inline `__oam_used:` label inside `IrOp::DrawSprite`
split the dead-load-elimination block, leaving the `STA $130 /
LDA $130` redundant store+load pair that main's peephole would
otherwise have collapsed to a plain `LDA #imm`. The stale bytes
shifted the NMI handler by a few bytes, which shifted `on frame`
execution enough that `examples/palette_and_background.ne` captured
phase 1 (WarmReds) at frame 180 instead of phase 2 (CoolBlues).
Regenerates every example ROM against the new codegen (all gate
behaviour is unchanged — the linker still sees the same markers,
just at the tail of the user stream instead of interleaved) and
updates the goldens that shifted: seven audio-hash drifts (all
audio-bearing programs, same cycle-accurate-APU-timing story as
every prior NMI layout change) and two pixel goldens — the one-
pixel sprite-position drift in `comparisons.png` that we already
tolerate, plus the phase-capture flip in
`palette_and_background.png`.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
* Tighten the OAM-DMA gate: the `has_oam` flag now ORs
`__sprite_cycle_used` in addition to `__oam_used`. A hypothetical
program that calls `cycle_sprites` without ever drawing would
otherwise compile to an NMI that advances \$07EF each frame but
never actually runs the DMA the cycling is meant to perturb.
The stronger gate keeps the two markers semantically coupled
(cycling presupposes DMA) and adds a test that verifies the
DMA trigger is emitted for a cycle-only program.
* Drop the unused `NmiOptions::any_input()` helper. The only
consumer (`gen_nmi`) reads the two flags inline and I never
wired up a second caller.
* Fix the cycle-count claim in `NmiOptions::has_p2_input` /
`has_p1_input` docstrings: LDA abs + LSR A + ROL zp is 11
cycles per port, ×8 loop iterations = ~88 cycles, not the "~30"
I wrote in the original commit. Also notes the ~12-cycle strobe
+ scaffold overhead that disappears when both ports are unused.
No behavioural change — all 622 Rust tests and 33 emulator
goldens still pass unchanged.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
With `has_p1_input` false, drop the three-instruction JOY1 shift
block from the NMI's input loop. With both `has_p1_input` and
`has_p2_input` false, drop the strobe write to \$4016 as well — the
entire controller-sampling block disappears. Audio- or compute-only
programs that never touch `button.*` pay zero cycles for input
sampling.
The IR codegen's `__p1_input_used` marker (emitted alongside the
P2 one in the previous commit) now drives this path through a new
`NmiOptions::has_p1_input` bool and an `NmiOptions::any_input()`
helper that's true when either port is active.
Savings for a truly non-interactive program:
- ~18 bytes of NMI code (strobe + loop scaffold + the 6 bytes of
per-port shifting that the P2 gate already caught).
- ~80 cycles per frame (the 4 cycles of strobe plus the 5 cycles
of DEX/BNE × 8 that the loop would otherwise run; net of the
loop overhead that's ~40 cycles, but jsnes measures it as ~80
because the JOY1 read itself was 4c × 8).
Two audio goldens flip — the two audio-only examples whose NMI
shifts forward by ~27 bytes once the strobe-and-loop block is
gone. Same cycle-accurate-APU-timing drift as every prior NMI
layout change.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Drop the three-instruction JOY2 shift block (`LDA $4017 / LSR A /
ROL ZP_INPUT_P2`) from inside the NMI's 8-iteration input loop
when user code never reads controller 2. IR codegen emits the
`__p2_input_used` marker from `IrOp::ReadInput(_, 1)`; the linker
threads the flag through a new `NmiOptions::has_p2_input` bool,
and `gen_nmi` writes the shift block only when the flag is set.
Savings for single-player programs:
- ~6 bytes of NMI code.
- ~30 cycles per frame (3 instructions × 8 loop iterations, each
6-8 cycles depending on addressing — LDA abs is 4, LSR A is 2,
ROL zp is 5, so ~11 cycles × 8 = ~88 cycles; rounded down for
the page-crossing penalty landing differently in the new layout).
This commit also fixes the IR codegen to drop the matching
`__p1_input_used` marker from `IrOp::ReadInput(_, 0)`, even though
the next commit is the one that actually consumes it. Landing the
two markers together keeps the IR codegen's per-op bookkeeping
coherent.
Six audio goldens flip (every program that reads input + plays
audio) with the expected NMI-layout-shift cycle drift.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Skip the OAM DMA (LDA#0/STA \$2003 + LDA#2/STA \$4014) inside the
NMI handler and the `\$FE` hide-sentinel fill of the \$0200 OAM
shadow inside `gen_init` for programs that never `draw`. Both are
gated on the `__oam_used` marker the IR codegen now drops at the
first `IrOp::DrawSprite`.
Savings per NMI for a non-drawing program:
- ~520 cycles (the DMA is 513 cycles plus the 4 register writes),
- ~9 bytes of NMI code,
- ~4 bytes of init code (the \$FE swap is replaced by a plain
zero-fill of \$0200-\$02FF alongside the rest of the 2 KB RAM
clear).
Plumbed by:
- New `NmiOptions::has_oam: bool`, threaded through `gen_nmi`.
- `gen_init(has_oam: bool)` parameter controlling the inner-loop
OAM fill. Existing runtime tests all migrate to `gen_init(true)`
to preserve their legacy assertions.
- Linker computes `has_oam = has_label(user_code, "__oam_used")`
once and feeds it to both call sites, and the existing
`has_visual_output` predicate reuses the same lookup rather than
re-scanning user_code.
sfx_pitch_envelope is the one audio-only example; its audio
golden flips by the usual cycle-accurate-APU-register-write-timing
drift caused by the NMI layout shifting ~14 bytes earlier.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Drop the built-in smiley from CHR tile 0 unless something in the
program actually references it. The marker fires when either:
1. `IrOp::DrawSprite` lowering falls back to tile 0 because the
sprite name doesn't resolve to a user declaration, or
2. The same lowering sees a runtime `frame:` override (which
could index any tile, including 0).
A third source of dependency — a background nametable entry of 0 —
is detected in the linker by scanning `bg.tiles` for zeros. This
preserves the smiley for programs like `examples/friendly_assets`
that use tile 0 as a background placeholder, even though their
draws resolve to user-declared sprites.
Programs whose draws all resolve to explicitly-declared sprites
with static frames AND whose backgrounds reference tiles 1+ now
leave CHR tile 0 as an all-zero blank, freeing 16 CHR bytes that
the user can treat as an always-transparent background tile.
Verified against the current example set: `sprites_and_palettes`
and `auto_chr_background` reclaim tile 0; every other example
keeps it (either they fall back to tile 0 via an undeclared draw
name or their background tilemap references tile 0).
All 33 emulator goldens still pass — removing an unreferenced CHR
tile can't change observable output.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Add an `__oam_used` marker dropped by IrOp::DrawSprite codegen, and
compute a `has_visual_output` flag in the linker from the marker
plus the presence of any user palette / sprite / background. When
that flag is false — i.e. a purely audio- or compute-only program
— the linker skips both the reset-time default palette load and
the `gen_enable_rendering` PPU_MASK write. `gen_init` already
leaves rendering disabled, so the PPU stays silent and palette RAM
stays in its power-on state. ~72 bytes reclaimed for non-visual
programs.
Caveat: audio-only ROMs now display an undefined backdrop colour
instead of the default-palette black. jsnes renders that as a
mid-grey; Mesen/real hardware may vary. Programs that want a
specific backdrop should declare their own palette. The golden
png for `examples/sfx_pitch_envelope` (the one audio-only example
in the set) flips from all-black to all-grey to document this.
`__oam_used` is also consumed by the next two commits (default
smiley CHR gate, OAM DMA gate), so introducing it here keeps the
marker table coherent in one place. Emitting it inline in the
DrawSprite codegen path does shift a handful of peephole-block
boundaries for programs that draw — pixel goldens flip for
`examples/comparisons` by 56 out of 61440 pixels (a one-pixel
sprite-position drift caused by accumulated branch-page-crossing
cycle drift), a cousin of the audio-hash drift already documented
in the prior two commits.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
The reset-time "no user palette" path was emitting 32 unrolled
`LDA #imm / STA $2007` pairs (~170 bytes) to write the built-in
palette. Replace it with the same indirect-loop loader the
user-palette path already uses (runtime::gen_initial_palette_load),
with the 32-byte default palette spliced into PRG under a
`__default_palette` data block. Net saving is ~120 bytes — ~20
bytes of code + 32 bytes of data vs ~170 bytes of unrolled stores.
Delete `Linker::gen_palette_load` (dead after the refactor) and its
unit test. Replace with two tests covering the observable
behaviour: the default palette bytes appear in PRG when no user
palette is declared, and the `__default_palette` label is
suppressed when the user does declare a palette.
Audio goldens flip again for audio_demo, noise_triangle_sfx, and
sfx_pitch_envelope. These are the three audio examples that don't
declare their own palette — shrinking the default-palette load
shifts their audio tick's absolute address by ~120 bytes, which
changes branch page-crossing timing and therefore the exact APU
register write sample offsets. Same class of drift as the
mul/divide gating commit.
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Drop __mul_used from IrOp::Mul codegen and __div_used from IrOp::Div
/ IrOp::Mod codegen (modulo reuses the same routine). The linker
skips gen_multiply / gen_divide for programs that never emit the
markers, following the same pattern already used by __audio_used /
__ppu_update_used / __sprite_cycle_used.
The optimizer already rewrites multiplies and divides by constant
powers of two into shifts (and modulo by constant powers of two
into masks), so the markers only fire for genuinely runtime math.
A program like `examples/comparisons.ne` that never multiplies or
divides now reclaims ~56 bytes of PRG; programs that use only one
of the two reclaim the other's share.
Audio goldens flip for every example that uses audio. The .ne
sources are unchanged and the pixel goldens are byte-identical —
the audio stream differs only because removing the math routines
shifts the audio tick's absolute address in PRG by 56 bytes, which
changes which of its internal branches cross 6502 page boundaries
and therefore the per-frame cycle count of a single NMI by 1-5
clocks. Over 180 frames the accumulated drift shifts APU register
write timing enough to render a different digital sample stream
at the same logical wave shape. Expected consequence of ROM-layout
change under cycle-accurate emulation; documented path per
CLAUDE.md "Updating goldens".
https://claude.ai/code/session_016kM6P7PukktBDqTZexrrAN
Three follow-ups from a fresh review of the perf milestone:
1. **UTF-8 safety in `substitute_asm_vars` and
`substitute_inline_const_params`.** Both walked the asm body
byte-by-byte and emitted each non-substituted byte via
`out.push(bytes[i] as char)` — a Latin-1 reinterpretation that
mangles non-ASCII characters in inline-asm comments. The brace-
level scan stays byte-based (braces can't appear inside a UTF-8
continuation), but the verbatim copy now uses
`out.push_str(&body[i..i + ch_len])` with `ch_len` derived from
the lead byte. Pre-existing latent bug in `substitute_asm_vars`,
freshly introduced in `substitute_inline_const_params` —
fixed in both, with a shared lead-byte length helper.
2. **`function_is_leaf` is now exhaustive on `IrOp`.** The match
used to be selective: `Call`/`Mul`/`Div`/`Mod`/`Transition`/
`InlineAsm` were checked, everything else fell through with
`_ => {}`. A new variant added later that secretly emitted a
JSR (e.g. a future `Mul16` calling `__multiply16`) would have
silently broken any leaf function that touched it. Listed
every current variant explicitly so the compiler errors at
the match arm if a new variant ships, and added a
`function_is_leaf_detects_jsr_emitting_ops` test that walks
the known JSR-emitting constructs (Call, *, /, %, asm with
JSR token) and asserts each disqualifies leafness.
3. **Cleanups.** `gen_block` now binds the fused-cmp dest temp
inside the original tuple instead of re-matching
`block.ops.last().unwrap()` to retire it. New
`inline_fun_with_asm_param_cascades_through_nested_inline`
test exercises the eval_const → const_args_stack path that
lets the inner of two nested inline funs see its outer's
parameter as the constant the top-level call passed. Defensive
comment on `body_has_inline_asm` explaining why it deliberately
doesn't recurse (relies on `is_splicable_void_stmt`'s
no-control-flow guarantee).
ROMs and goldens unchanged — all the changes are non-observable
through the existing example surface. Verified: cargo
test/clippy/fmt clean on rustc 1.95.0; emulator harness 34/34;
reproducibility diff clean; demo gifs byte-match fresh captures.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
`inline fun` worked for plain-NEScript bodies but blew up on any
function that used inline asm with `{name}` substitution. The
codegen's `substitute_asm_vars` runs against the analyzer's
per-function scope; after splicing, the scope is the *caller*,
where the inlined fun's parameters don't exist. The result was
`{dst}` left as a literal token in the spliced asm body, and
the asm parser failing with `bad number `{dst}``.
Fix: at inline-expansion time (`try_inline_call_stmt`), when the
splicer detects a `Statement::InlineAsm` in the body and the
call site passed a compile-time constant for a parameter,
pre-substitute `{param}` with `#$<value>` so the spliced body
parses as an immediate-mode operand. The substitution is done
via a new `inline_const_args_stack` parallel to the existing
`inline_subs_stack`, populated from the args' `eval_const`
results.
When *any* arg is non-constant the splicer refuses to inline an
asm-containing function and falls back to a regular `Call` op —
preserving correctness, just at the cost of the JSR/RTS apparatus
the user was hoping to avoid. The fallback is exercised by the
new `inline_fun_with_asm_falls_back_for_runtime_arg` test.
`eval_const` is also extended to consult the same const-args
stack, so a nested inline like `inline_outer(K)` →
`inner(outer_param)` correctly recognises `outer_param` as the
constant `K` and recurses the substitution. Without this, the
cascade stopped at the first level.
Two new tests in `src/ir/tests.rs` lock in the behaviour:
- `inline_fun_with_asm_param_substitutes_immediate` — verifies
`{param}` becomes `#$<value>` in the spliced body and no
Call op is left.
- `inline_fun_with_asm_falls_back_for_runtime_arg` — verifies
the fallback path emits a Call op.
The SHA-256 example doesn't itself opt into the new feature for
its primitives (full inlining would balloon ROM by 5-10 KB —
the call sites add up fast at 1500+ source-level uses); that's
left as a future opportunity. Hash output unchanged; emulator
harness 34/34; reproducibility diff clean.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
The IR codegen lowers `i -= 1` (and friends) into a `LoadImm temp,
1; Sub d, i, temp; StoreVar i, d` triple, and the optimizer
strength-reduces the Sub+StoreVar pair into `DEC i`. The
constant-load-into-A that used to feed the Sub stays around as a
dead `LDA #1`:
LDA #1
DEC ZeroPage(rem)
JMP Label("__ir_blk_while_cond_…")
`remove_dead_loads` was set up to drop exactly this pattern but
gave up at the trailing `JMP` because it couldn't reason about
flow. Extend it to follow one unconditional `JMP <label>` to its
target and resume the dead-store scan from the next instruction.
The first instruction past the loop-condition label is reliably an
`LDA loop_var`, which overwrites A without reading it — so the
`LDA #1` is correctly identified as dead.
Conditional branches still end the scan (their not-taken path is
unconstrained) and only one JMP is followed (to keep the analysis
local). For SHA-256 specifically this drops two `LDA #1`s per
iteration of the rotate/shift bit-loops — about 1K cycles per
block. The same pattern fires across most examples' loop tails.
Verified: cargo test/clippy/fmt clean on rustc 1.95.0; emulator
harness 34/34; reproducibility diff clean; SHA-256 of "NES" still
computes to AE9145DB…4E0D. The cycle drift refreshes the four
audio hashes / golden frames timing-sensitive examples already
tracked.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
Every NEScript condition (`if x < N`, `while i < end`, etc.)
lowers in two IR ops: `CmpX(d, a, b)` materializes a 0/1
boolean into temp `d`, and the block's terminator
`Branch(d, t, f)` reads `d` and branches on it. The codegen
faithfully emitted both halves — `LDA / CMP / branch-to-true /
LDA #0 / JMP done / true: LDA #1 / done:`, then later
`LDA d_slot / BNE branch_t / JMP branch_f` — about 14 cycles +
13 bytes per condition.
The 6502's natural pattern is one `CMP` + one branch on the
flags it just set: 8 cycles, no register-clobber, no temp slot.
Detect the canonical pattern in `gen_block` (last op is an 8-bit
`CmpX` whose dest temp is what the terminator branches on, with
no other uses) and emit the fused form directly via a new
`gen_cmp_branch` helper. The temp's allocation, store, load, and
the terminator's branch fall away.
Bookkeeping subtlety: the source temps `a`/`b` must be retired
*after* the fused emit, not before — the original `gen_op` order
is "emit body of op, then `retire_op_sources`". Decrementing
their use counts before the CMP would free their slots while
they were still live; `load_temp(a)` would then re-allocate `a`
to whatever stale slot the free list popped next. Got hit by
this on the first attempt — the SHA-256 example dutifully
returned all-zero hashes until the order was fixed.
Updated `ir_codegen_local_label_suffix_is_bank_namespaced`: the
test was relying on `if x == 0` to emit `__ir_cmp_*` labels for
its bank-namespacing check, which the fusion now collapses into
direct branches. Switched the test source to a shift-by-variable
pattern (`x = x << n`), which always emits `__ir_shift_loop_*`
labels regardless of future cmp/branch optimizations.
Cycle savings: ~6 cycles per condition. The SHA-256 rotate
loops alone account for ~9K cycles per block. Across all
examples the cycle drift shows up as audio-tick phase shifts
in five timing-sensitive ROMs (`audio_demo`, `friendly_assets`,
`noise_triangle_sfx`, `platformer`, `sfx_pitch_envelope`); the
goldens for those are refreshed in this commit, plus
`platformer.gif` (the only demo gif whose bytes actually moved).
Verified: cargo test/clippy/fmt clean on rustc 1.95.0;
emulator harness 34/34; reproducibility diff clean; SHA-256 of
"NES" still computes to AE9145DB…4E0D.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
Leaf functions — those that never JSR another routine from inside
their body — don't need to spill the `$04..$07` parameter
transport slots into per-function RAM, because nothing inside the
body clobbers those slots. Detect them in `IrCodeGen::new` via a
linear scan over each function's IR ops, point their parameters
at `$04..$07` directly in `var_addrs` (and in a parallel
`leaf_param_overrides` map for inline-asm `{name}` substitution),
and have `gen_function` skip the spill prologue.
The "leaf" predicate is conservative: any of `IrOp::Call`, `Mul`,
`Div`, `Mod`, `Transition`, or an inline-asm body containing a
`JSR` token disqualifies the function. SetPalette /
LoadBackground / PlaySfx / StartMusic / DebugLog / DebugAssert
were verified by inspection to not emit JSRs.
Per call to a leaf primitive: `LDA $04 / STA <local> / LDA $05 /
STA <local+1>` is now omitted — saves 12 cycles and 12 bytes of
code per call. Across the SHA-256 example's ~5500 leaf-primitive
calls per block, that's ~66K cycles saved per compression — about
2.2 frames at NTSC.
The fix also touches every committed `examples/*.nes` (the leaf
prologue was emitted by every fun with params, not just the SHA
ones), so 9 ROMs and the same three timing-sensitive goldens
(war.png + platformer/pong/war audio hashes) get refreshed; the
two committed gifs that drifted do too.
Verified: cargo test/clippy/fmt clean on rustc 1.95.0; emulator
harness 34/34; reproducibility diff clean; SHA-256 of "NES" still
computes to AE9145DB…4E0D.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
Fixes compiler-bugs.md #1 — the inline-asm `{name}` resolver
looks parameters up in the analyzer's `VarAllocation` table
(because that's the only address map it has), but `IrCodeGen::new`
was minting a parallel `$0300+` range for every function-local and
ignoring what the analyzer had picked. The spill prologue wrote the
param to the codegen's private address, the inline asm read from
the analyzer's zero-page address, and nothing ever bridged the two
— `LDA {param}` would silently load whatever the RAM clear left at
the stale slot (always `0`).
Fix: drop the `local_ram_next` loop and just look each local up in
`allocations` by the analyzer's qualified name
(`__local__{scope}__{local}`). The scope string that `gen_function`
already computed for `substitute_asm_vars` is now shared with the
new address-seeding loop via a `scope_prefix_for_fn(&str)` helper,
so the two call sites can't drift. The analyzer's layout already
satisfies the "no overlapping live locals" invariant the codegen
was relying on — it scopes every local under
`__local__<scope>__<name>` so two functions with a parameter named
`x` land in different slots.
Updated `gen_function_prologue_spills_params_to_local_ram`: the
regression test for the War-era param clobbering bug was asserting
the spill's destination specifically had to be an absolute address
at `$0300+`. That's no longer the mechanism — the spill lands in
whatever slot the analyzer assigned, which is zero page when
there's room. The test now asserts the destination is *any*
address outside `$04-$07`, which is the actual invariant.
Reverted the `LDX $04` / `LDY $05` workaround in
`examples/sha256/sha_core.ne` — every primitive there now uses
`{dst}` / `{src}` / `{w_ofs}` / `{h_ofs}` / `{k_ofs}` substitution
as originally intended. The "Parameter convention" comment that
documented the workaround is gone.
Regenerated `tests/emulator/goldens/inline_asm_demo.png`: that
example's `times_four(input)` was previously returning `input`
verbatim because the inline asm's `LDA {result}` / `ASL A` /
`ASL A` / `STA {result}` operated on a zero-page byte that was
disconnected from the NEScript-level `result` variable. With the
fix, `times_four` correctly returns `input * 4`, so the
smiley-tracker's frame-180 position shifts by the expected
`(frame_count * 4) mod 256` delta. The other 33 ROMs remain
byte-identical.
Verified:
- `cargo clippy --all-targets -- -D warnings` clean on both
rustc 1.94.1 and 1.95.0.
- `cargo test --all-targets`: 616 + 3 + 75 tests pass.
- `cargo fmt --check` clean.
- Full emulator harness: 34/34 ROMs match goldens.
- SHA-256 of "NES" still computes to
`AE9145DB5CABC41FE34B54E34AF8881F462362EA20FD8F861B26532FFBB84E0D`.
- `--memory-map` output now reflects what the generated code
actually reads and writes (previously the codegen's $0300+
override was invisible to the dump).
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
Rust stable rolled to 1.95.0 today (2026-04-16), which fires four
new warning-by-default lints on pre-existing code. All four have
mechanical fixes suggested by clippy itself:
- `collapsible_match` (2x) in `src/analyzer/mod.rs` — merge the
`if args.len() != N` guard into the `match` arm pattern. One
diagnostic-push per arm, shape is identical to before.
- `unnecessary_sort_by` in `src/optimizer/mod.rs` — replace
`sort_by(|a, b| b.1.cmp(&a.1))` with
`sort_by_key(|(_, c)| std::cmp::Reverse(*c))`.
- `manual_checked_ops` in `src/optimizer/mod.rs::Div` folding —
replace `if vb == 0 { 0 } else { va / vb }` with
`va.checked_div(vb).unwrap_or(0)`. Same `x / 0 == 0` fallback.
- `map_unwrap_or` in `src/main.rs` — `std::fs::metadata(...).
map(|m| m.len()).unwrap_or(0)` → `.map_or(0, |m| m.len())`.
Verified with both the old 1.94.1 stable and the freshly-installed
1.95.0: `cargo clippy --all-targets -- -D warnings` is clean on
both; `cargo fmt --check` is clean; `cargo test --all-targets`
still passes 616 + 3 + 75; the emulator harness still matches
all 34 goldens byte-for-byte.
This unblocks the SHA-256 PR (imjasonh/nescript#28) and any
other PRs that run CI against Rust 1.95.
https://claude.ai/code/session_01FRmSBruVWCufm3LsUVMs8v
W0110: when a function marked `inline` has a body shape the IR
lowerer can't splice (conditional early return, loops, nested
control flow, empty void body), the analyzer now emits a
warning at the declaration site so the declined hint is
visible instead of silently falling back to a regular JSR.
Implementation:
- New `W0110` error code in `src/errors/diagnostic.rs` (warning level).
- New `pub fn can_inline_fun(return_type, body) -> bool` in
`src/ir/lowering.rs`, extracted from the existing capture
logic so the analyzer and the IR lowerer share the same
eligibility rules and can never drift.
- New `check_inline_declinability` analyzer pass called from
the tail of `analyze_program`, mirroring the existing
`check_sprite_scanline_budget` / `check_unreachable_states`
passes. Emits W0110 with help + note text pointing at the
two accepted body shapes.
- `capture_inline_bodies` now defers to `can_inline_fun`
instead of duplicating the match pattern, so the two sides
stay in lockstep by construction.
Four regression tests in `src/analyzer/tests.rs` cover the
conditional-return and while-loop declines plus the two
accepted shapes (single-return expression, void sequence).
Example source cleanups: `wrap52` in `examples/war/deck.ne`
and `abs_diff` in both `examples/arrays_and_functions.ne` and
`examples/loop_break_continue.ne` drop the `inline` keyword.
All three were dead hints — the `inline` was being silently
declined before this change, so removing it is source-only;
the three ROMs are byte-identical, all 32 emulator goldens
still match.
Docs refresh
- `docs/language-guide.md`: rewrote the Inline Functions section
(real behaviour + W0110), added W0105/W0106/W0107/W0108/W0109/
W0110 to the warnings table, added the `debug.sprite_overflow*`
builtins + sprite-per-scanline mitigations section to the
Debug Mode docs, added a `cycle_sprites` statement entry and
cross-referenced it from `draw`.
- `docs/nes-reference.md`: fleshed out the "NEScript Memory
Usage" block with the full ZP + high-RAM layout, including
the new `$07EF` / `$07FC` / `$07FD` slots for sprite cycling
and the debug sprite-overflow telemetry.
- `docs/future-work.md`: documented all four debug query
builtins in the "What ships today" block; updated the open
"OAM allocation strategy" question to reference the shipped
`cycle_sprites` path and ask about an automatic-flicker
game attribute as a follow-up.
- `docs/architecture.md`: updated the `ir/` and `optimizer/`
module summaries to describe real inline splicing (now
in lowering, not the optimizer).
- `README.md`: reframed the `inline` bullet from "hint" to
"real splicing for single-return / void-body shapes";
expanded the debug-support bullet to mention the four
query builtins and their stripping in release builds; added
a new bullet for the three-layer sprite-per-scanline
mitigations; bumped the test count from 497 → 694; updated
the war.ne entry to mention the seven compiler bugs are all
fixed and point readers at `git log` (instead of the
deleted COMPILER_BUGS.md).
- `examples/README.md`: same `git log`-pointing rewrite for
the war.ne entry.
Deletions
- `examples/war/COMPILER_BUGS.md` is removed. All seven
catalogued bugs are fixed; the file's historical value
lives in `git log` now. Every source-code comment and doc
reference to the file has been updated to either point at
`git log` or just describe the bug in place.
Test count: 616 unit + 75 integration + 3 doctests = 694 total.
Clippy / fmt clean. 32/32 emulator goldens match.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
W0109 (shipped last commit) catches the 8-sprites-per-scanline
hardware limit at compile time for static layouts, but the
dynamic case — enemy formations, projectile clusters, animated
NPCs where coordinates come from variables — was still silent.
This change adds two layers of defense on top of W0109:
Layer 2: `cycle_sprites` runtime flicker intrinsic
New keyword statement that rotates the OAM DMA start offset
one slot per call. When called once per `on frame`, the PPU's
sprite evaluation picks up a different subset of the 12+
overlapping sprites each frame, so the permanent-dropout
failure mode becomes visible flicker — the classic NES
technique used by Gradius, Battletoads, and every shmup.
Implementation:
- Lexer keyword `KwCycleSprites` and parser production.
- AST `Statement::CycleSprites(Span)`.
- `IrOp::CycleSprites` lowered by the IR pass.
- Codegen emits `LDA $07EF / CLC / ADC #4 / STA $07EF` with
natural u8 wrap, plus a one-shot `__sprite_cycle_used`
marker label the first time it fires.
- Linker detects the marker and switches `gen_nmi` to the
cycling variant, which reads the rotating offset from
`$07EF` into OAM_ADDR before the DMA instead of writing
a literal 0. Programs that don't call `cycle_sprites`
skip the marker and get byte-identical ROM output.
Layer 3: debug-mode sprite overflow telemetry
Mirrors the frame-overrun pair (`debug.frame_overrun_count` /
`debug.frame_overran`). In debug builds the NMI handler reads
`$2002` at the top of vblank, masks bit 5 (the PPU's sprite
overflow flag), and if set bumps a cumulative counter at
`$07FD` plus a sticky bit at `$07FC`. The sticky bit clears
on every `wait_frame`.
New debug builtins:
- `debug.sprite_overflow_count()` → u8 peek of $07FD
- `debug.sprite_overflow()` → u8 peek of $07FC (sticky bit)
The hardware flag has well-known quirks but is correct for
the overwhelming majority of cases and costs ~15 cycles per
frame to sample. Release builds emit no overflow-check code
at all, so the four bytes at `$07EF` / `$07FC`-`$07FD` stay
free for user allocation.
Related changes:
- `gen_nmi` now takes an `NmiOptions` struct. Four bool
parameters tripped clippy's `fn_params_excessive_bools`.
- CLI `build` now renders analyzer warnings on a successful
build. Previously warnings were silently dropped unless
the user also ran `nescript check`, which made W0109
effectively invisible to CI and local dev alike. Existing
pre-existing W0103 / W0106 warnings on `coin_cavern`,
`mmc3_per_state_split`, `sprites_and_palettes` surface
too — not regressions, just now visible.
New example: `examples/sprite_flicker_demo.ne`
Draws 12 sprites into a 4-pixel band, W0109 fires at compile
time with nine labels pointing at the offenders, and a
`cycle_sprites` call at the end of `on frame` turns the
hardware dropout into flicker. The committed emulator golden
captures one frame of the cycling pattern (deterministic).
Tests:
- `runtime::tests::nmi_debug_mode_samples_sprite_overflow`
- `runtime::tests::nmi_sprite_cycle_variant_reads_rotating_offset`
- `ir_codegen::*::debug_sprite_overflow_count_loads_07fd`
- `ir_codegen::*::debug_sprite_overflow_flag_loads_07fc`
- `ir_codegen::*::wait_frame_clears_sprite_overflow_sticky_in_debug_mode`
- `ir_codegen::*::wait_frame_release_does_not_touch_sprite_overflow_sticky`
- `ir_codegen::*::cycle_sprites_emits_marker_and_add4`
- `ir_codegen::*::cycle_sprites_marker_dedup_across_multiple_calls`
- `ir_codegen::*::program_without_cycle_sprites_emits_no_marker`
- `analyzer::*::accepts_debug_sprite_overflow_builtins`
- `analyzer::*::rejects_unknown_debug_method_lists_all_four_known_names`
- `analyzer::*::accepts_cycle_sprites_statement`
Docs: `examples/war/COMPILER_BUGS.md` §4 now describes all three
layers (W0109, `cycle_sprites`, debug telemetry) with reasoning
for when each applies. `README.md` and `examples/README.md` add
the new example to their tables.
All 32 emulator goldens still match — the cycling is opt-in
and programs that don't call `cycle_sprites` or enable debug
mode are byte-identical to the pre-change output.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
Fixes the last two deferred compiler bugs catalogued in
examples/war/COMPILER_BUGS.md, finishing the bug-cleanup arc on
the War branch.
Bug #5 — `inline fun` inliner
Previously the `inline` keyword was parsed into `FunDecl.is_inline`
and then dropped on the floor: every call site emitted a regular
`JSR` through the $04-$07 transport slots. Now the IR lowerer
captures inline function bodies up front in
`LoweringContext::capture_inline_bodies` and rewrites call sites
at lowering time. Two body shapes are supported:
1. Single-return expression — the body is re-lowered in place
of the `Call` op with the parameter names substituted to
fresh IR temps for each argument.
2. Void multi-statement body whose every statement is one of
Assign/Call/Draw/Scroll/SetPalette/LoadBackground/WaitFrame/
Play/StartMusic/StopMusic/InlineAsm/RawAsm/DebugLog/DebugAssert
— the statements are spliced into the caller's block with
the same parameter substitution machinery.
Control-flow-heavy inline bodies (conditional early returns,
loops, transitions) fall back to a regular out-of-line call with
no diagnostic. That's predictable and documented in the bug-tracking
doc. Nested inline expansion uses a substitution-frame stack so
an inline calling another inline sees the right arguments.
A codegen follow-up was needed because bug #3's scope-qualified
local names broke `{result}` substitution in inline asm. The
codegen now tracks `current_fn_scope_prefix` per function and the
InlineAsm op tries the qualified name first before falling back
to the bare name.
Bug #4 — W0109 sprite-per-scanline static check
Adds a new warning code W0109 and an analyzer pass
`check_sprite_scanline_budget` that walks each state's `on_frame`
handler, collects literal-coordinate `draw` statements (including
metasprite expansion via dx/dy offsets), and iterates scanlines
0..240 to count how many 8x8 sprites overlap each line. When a
scanline has > 8, the analyzer emits W0109 with labels pointing
at each offending draw site plus a help message about staggering
y-rows and a note explaining the hardware dropout. Non-literal
coordinates are skipped (static analysis can't resolve them).
Nested `if`/`while`/`for`/`loop` blocks are unioned conservatively.
Tests added
src/ir/tests.rs
- inline_fun_expression_body_emits_no_call_at_use_site
- inline_fun_void_body_statements_are_spliced
- inline_fun_with_conditional_return_compiles_as_regular_call
- inline_fun_nested_inlines_substitute_correctly
src/analyzer/tests.rs
- analyze_sprite_scanline_budget_warns_over_eight
- analyze_sprite_scanline_budget_ok_when_staggered
- analyze_sprite_scanline_budget_skips_dynamic_coords
- analyze_sprite_scanline_budget_expands_metasprites
- analyze_sprite_scanline_budget_recurses_into_if
COMPILER_BUGS.md
Bugs #4 and #5 marked **FIXED** in the status table, with full
reproduction/root-cause/fix/regression-test write-ups updated in
place. All seven catalogued bugs now have shipped fixes.
Artifact churn
- examples/war.nes and examples/inline_asm_demo.nes rebuild
byte-shifted (different JSR targets post-inliner).
- tests/emulator/goldens/war.audio.hash shifts from 143660f to
13443e28 — the inliner removes JSRs to set_phase, which nudges
NMI sampling timing. No pixel diff; behavior is unchanged.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
Three related scoping bugs from examples/war/COMPILER_BUGS.md,
all fixed in one pass because they're different layer
manifestations of the same "flat global namespace" problem:
## §3: function-local `var` declarations lived in one namespace
`src/analyzer/mod.rs::register_var` inserted every `var` it
saw — top-level, state-local, AND function-body local — into
the same `self.symbols: HashMap<String, Symbol>`. Two different
functions declaring `var i` collided on E0501, which is why
every local in war/*.ne had a function-prefix like `dfa_card`
or `dwp_px`.
Fix: add a `current_scope_prefix: Option<String>` to the
Analyzer, set it to `Some("<fn_name>")` when checking a
function body (or `Some("Title__frame")` for state handler
bodies), and have `register_var` store the declaration under
an internal key `"__local__{prefix}__{name}"`. New
`resolve_symbol` / `resolve_key` helpers try the
scope-qualified key first and fall back to the bare key for
globals / consts / enum variants / state-level vars / function
names. Every existing `self.symbols.get(name)` inside
body-checking code was swapped over.
Two `var i` declarations inside the SAME function body still
collide with E0501 — we scoped per function body, not per
nested block. Per-block scoping would require live-range
analysis to reuse RAM slots.
## §1b: same-named params across functions shared VarIds
`src/ir/lowering.rs::get_or_create_var` looked up names in a
single global `var_map`, so two functions both with a `card:
u8` parameter resolved to the same `VarId`. Whichever function
was lowered last won the zero-page slot mapping, silently
rerouting the other function's param reads to the wrong slot.
Fix: the IR lowerer now mirrors the analyzer's scope logic.
`LoweringContext` gains a `current_scope_prefix` field that
gets set in `lower_function` / `lower_handler`, and
`get_or_create_var` uses a new `scoped_key` helper that
prepends `"__local__{prefix}__"` when the qualified key exists
in `var_map` or `var_types`. Each function's parameters and
locals therefore get distinct VarIds, and the codegen's
`var_addrs` map naturally has no collisions.
## §2: param transport slots $04-$07 clobbered across nested JSRs
Parameters were passed AND kept in `$04-$07` for the lifetime
of a function. Any nested call overwrote those slots with its
own arguments, so the caller's params were silently corrupted
as soon as it invoked anything. Every war helper that took
params and called other helpers (draw_card_face, push_back_a,
etc) snapshotted its params into fresh locals at the top of
the body.
Fix: in `codegen/ir_codegen.rs::IrCodeGen::new`, every
function-local — including parameters — now gets a dedicated
per-function RAM slot at `$0300+`. Parameters are still passed
via the zero-page transport slots `$04-$07` as the calling
convention, but `gen_function` now emits a **prologue** at
every function entry:
LDA $04
STA <param_0_addr>
LDA $05
STA <param_1_addr>
... etc, up to 4 ...
By the time the body runs, every parameter lives in the
function's dedicated RAM slot, so any nested call can freely
clobber $04-$07 (writing its own arguments there) without
corrupting the caller's saved parameters. Costs 4 LDA/STA
pairs (≈ 20 bytes of ROM, 16 cycles) at every function entry
— worth it to make the calling convention sound.
## War cleanup
With all three fixes in place, every workaround prefix in
`examples/war/*.ne` is gone:
- `card_rank(card)` instead of `card_rank(crk_c)` — bug #1b
- `compare_cards(a, b)` instead of `compare_cards(cmp_a, cmp_b)`
- `push_back_a(card)` instead of `push_back_a(pba_in)` — bug #1b
- `var card: u8 = draw_front_a()` in bury_from_* — bug #3
- `var i: u8 = 0` freely in multiple functions — bug #3
- `fun push_back_a(card)` body no longer snapshots `card` into
`pba_card` before calling wrap52 — bug #2
- `fun draw_card_face` body no longer snapshots x/y/card into
locals before calling card_rank/card_suit — bug #2
- `draw_word_player` steps its own x without needing a
`dwp_px` accumulator to avoid the `x + N` arg compilation
quirk — that quirk was a downstream symptom of bug #2 and
is also gone
The source is now about 300 lines shorter and significantly
more readable.
## Regression tests
Seven new tests nail these bugs down:
- `analyzer::tests::analyze_allows_same_local_name_in_two_functions`
- `analyzer::tests::analyze_allows_same_param_name_in_two_functions`
- `analyzer::tests::analyze_allows_same_local_name_in_two_state_handlers`
- `analyzer::tests::analyze_still_rejects_duplicate_local_in_same_function`
- `codegen::ir_codegen::gen_function_prologue_spills_params_to_local_ram`
Plus the four param-arity tests from the earlier E0506 fix
and the wide_hi-leak regression test from the previous
compiler fix. Total suite: 591 unit tests, all passing.
## Golden drift
The prologue change adds a few cycles to every function entry,
which shifts NMI sampling by a handful of cycles and flips
the audio-hash of any example that plays sfx or music
(platformer, war). `arrays_and_functions.png` also picks up a
1-pixel shift in its enemy positions due to the same timing
drift. All three golden updates are pure "compiler produces
different but functionally-identical output" — no game
behavior changed.
## What's still open in COMPILER_BUGS.md
- §4: 8-sprites-per-scanline hardware limit is invisible to
user code. A static analyzer hint could help; deferred.
- §5: `inline` keyword is silently declined for short
functions that the optimizer's inliner doesn't recognize
(it only removes empty functions). Deferred pending a real
single-return-expression inlining pass.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
The IrLowerer's wide_hi map records "this u8 temp's high byte
lives at this other temp" pairs whenever a 16-bit value is
produced. Both lower_function and lower_handler reset next_temp
to 0 at the start of each function, but neither cleared wide_hi
— so stale (low_id -> high_id) entries from earlier functions
leaked into subsequent ones.
When a fresh function reused those temp IDs for unrelated u8
expressions, is_wide() returned spurious true and widen() handed
back stale (lo, hi) pairs whose hi happened to coincide with the
*next* temp ID fresh_temp() was about to allocate. The result
was 16-bit IR ops (CmpEq16 in particular) where the destination
temp aliased one of the source operand high bytes — for War this
made `match phase` arms past P_WIN_B impossible to enter and the
game would freeze with both face-up cards on the table forever.
Fix: clear wide_hi alongside the next_temp reset in both
lower_function and lower_handler. Adds a regression test
(ir::tests::wide_hi_does_not_leak_between_functions) that
constructs a function whose body has no u16 ops but follows a
function that does, and asserts no CmpEq16 op aliases its dest
with an operand high byte.
Also:
- Convert the war Playing state's phase machine from an
if-chain to a `match`, which is what tripped this bug to the
surface (it was lurking in earlier ROMs too but their layouts
never produced the dest/source collision shape).
- Refactor begin_draw_a/b to set fly_card / fly_face_up via
globals before calling arm_fly, since arm_fly only takes 4
params (the v0.1 ABI limit, now diagnosed by E0506).
- Hoist the P_RESOLVE comparison result to the global pf_result
to dodge the param-clobbering issue documented in
examples/war/COMPILER_BUGS.md §2.
- Document the bug as item #6 in COMPILER_BUGS.md with a
minimal repro and reproducer-test pointer.
- Refresh the war golden + audio hash to match the new ROM.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
The v0.1 calling convention passes parameters through four fixed
zero-page slots ($04-$07). Functions declared with 5+ parameters
were silently dropped past the 4th, producing a runtime miscompile
with no compile-time signal — a trap I hit while building the
War example (arm_fly took 6 params and silently corrupted fly_card
and fly_face_up).
Add E0506 to the analyzer so the over-arity case becomes a clear
compile-time error pointing at the user's `fun` declaration with
guidance toward globals or splitting. New tests cover both the 5-
param rejection and the 4-param accept boundary.
Documented in examples/war/COMPILER_BUGS.md §1, language-guide.md
"Restrictions" section, and the error code table.
https://claude.ai/code/session_0143dTgh3UeRrtfHgQwzcv5z
A focused review of the branch surfaced two correctness bugs and
four important polish items in the new features. None of the
existing example goldens shift — every fix is gated on conditions
that don't fire in the committed examples.
**debug.frame_overran() never reset between frames in implicit
wait_frame programs.** The IR-level WaitFrame op cleared $07FE,
but the implicit main-loop flag-clear that runs between dispatch
iterations only cleared ZP_FRAME_FLAG. A program whose
`on frame { ... }` body had no explicit `wait_frame` would latch
$07FE to 1 on the first miss and never reset, breaking
`debug.assert(not debug.frame_overran())` guards. The dispatch
loop now also clears $07FE in debug builds, mirroring the
WaitFrame path. New regression test asserts the main loop emits
exactly one STA $07FE in a no-wait_frame debug build.
**Metasprite base-tile resolution silently miscompiled for
`@chr` / `@binary` sprites.** The IR lowering walks
`program.sprites` to compute base tile indices but assumes
1 tile per non-Inline source, while the real asset resolver
reads the file. The analyzer now hard-rejects the combination
with a clear "use inline pixels" hint instead of letting it
compile to a visual glitch. New analyzer test
`analyze_metasprite_with_external_chr_sprite_errors` covers it.
**next_sprite_tile capping silently allowed CHR overlap.** The
pipeline used `.min(255)` which would let a background tile
overwrite a sprite tile when the sprite range filled the
pattern table. Now hard-errors via CompileError::AssetResolution
when the sprite range >= 256 *and* the program declares any
`@nametable(...)` background. Inline backgrounds aren't affected.
**Linker silently truncated background CHR overflow.** The
`if end <= chr.len()` guard at the CHR copy site dropped any
auto-CHR bytes that would have run past the pattern table.
Replaced with a debug assertion since the resolver should
have caught it upstream — defense in depth.
**Stale comment in nested_structs.ne** said struct literals
"don't accept array fields yet" while the example itself
demonstrates inline array fields working through
`expand_struct_literal_init`. Comment updated.
**Misleading sentinel comment in audio.rs** described the pitch
envelope's trailing zero as a runtime sentinel; in practice the
volume tick `JMP`s to `__audio_sfx_done` first and the pitch
update block never reads the trailing byte. Rewrote the comment
to clarify it's padding for predictable blob length.
Also tidies up two minor items the reviewer flagged:
- `flatten_struct_fields` rebuilt the `struct_sizes` HashMap on
every leaf field; hoisted the snapshot to the function entry.
- Integration tests called `resolve_backgrounds(..., 0)` (the new
`next_sprite_tile` parameter); changed to `1` so a future
PNG-nametable test fixture won't accidentally overwrite the
runtime smiley at tile 0.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
Adds a focused unit test that constructs a `SpriteData` at tile 1
and a `BackgroundData` whose `chr_bytes` claim tiles 5-6, then
verifies the linker's CHR ROM placement preserves the smiley at
tile 0, the sprite at tile 1, leaves tiles 2-4 untouched, and
copies the background blob at the requested base offset. Catches
any future regression in the
`BackgroundData::chr_base_tile` → CHR ROM splice that
`assets: auto-generate CHR data from @nametable() PNG sources`
introduced.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
`background Foo @nametable("file.png")` previously decoded the PNG
into a tile-index table and an attribute table but left CHR
generation to the user — they had to supply matching tiles via a
separate `sprite Tileset @chr(...)` declaration in the same
deduplication order, which was both error-prone and the main thing
keeping the shortcut form from being a one-liner.
The CHR pipeline now closes the gap. `png_to_nametable_with_chr`
returns a `PngNametable` carrying the tile-index table, the
attribute table, *and* a per-tile CHR blob encoded with the same
brightness-bucketing `png_to_chr` already uses for sprites. The
resolver passes `next_sprite_tile` (computed from the resolved
sprite list) so each background's CHR allocation slots in
immediately after the sprite range, and rewrites the nametable
indices to point at the actual physical tile numbers. The linker
copies each background's `chr_bytes` into CHR ROM at
`chr_base_tile * 16`, so the final image renders without any
user-supplied CHR.
`BackgroundData` carries `chr_bytes` and `chr_base_tile` so the
linker has everything it needs at a glance. Inline `tiles:` /
`attributes:` declarations leave them empty and behave exactly
like before — that path doesn't auto-generate CHR because the
user is implicitly opting into "I'll provide tiles myself" by
typing the indices out by hand.
The new `examples/auto_chr_background.ne` is a 256×240 grayscale
gradient committed alongside its `auto_chr_bg.png` source; the
emulator harness verifies the rendered output against a
committed golden so a regression in the dedupe/encode/linker
plumbing fails CI loudly. Existing example ROMs are byte-
identical because their backgrounds either have no PNG source or
already provided their own CHR.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
Multi-tile sprites used to require one hand-written `draw` per tile,
e.g. the four-call sequence in `examples/platformer.ne`'s
`draw_player()`. The new `metasprite Name { ... }` declaration
collects parallel `dx`/`dy`/`frame` arrays plus a reference to the
underlying sprite, and `draw Name at: (x, y)` expands to one OAM
slot per tile in the IR lowering — the codegen sees N regular
DrawSprite ops, so the runtime OAM cursor allocator picks them up
without any metasprite-specific awareness.
The metasprite's `frame:` array is interpreted *relative to the
underlying sprite's base tile*: index 0 means "the first tile this
sprite owns", which is the natural reading for a 16×16 hero whose
pixel art the asset resolver split into four consecutive tiles.
The lowering walks `program.sprites` to compute base tile indices
the same way `assets::resolve_sprites` would, then folds the base
into each frame entry before storing the metasprite info. Sprites
sourced from external `@chr(...)` / `@binary(...)` files whose
bytes aren't available at parse time fall back to a one-tile
assumption — those programs are rare and can declare metasprites
against pixel-art sprites instead.
The new `examples/metasprite_demo.ne` declares a 16×16 hero sprite
and arranges its four tiles into a metasprite, then sweeps the
hero across the screen so the harness captures it mid-motion.
The new keyword is added to the lexer/token list, and the parser
accepts `sprite:` (the otherwise-keyword) as a property name in
metasprite bodies so the natural spelling parses.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
Pulse-channel sfx with a multi-byte `pitch:` array used to silently
ignore everything past the first byte — the runtime audio tick
latched the period at trigger time and never updated it. Programs
that wanted a frequency sweep had no way to express it.
The compiler now compiles a per-frame pitch envelope blob alongside
the existing volume envelope when `decl.pitch` has more than one
distinct value. The blob is padded (or truncated) to the volume
envelope's length and ends in a zero sentinel so the runtime
walker stops both pointers on the same NMI. Sfx with a single
scalar pitch (or an array where every byte is the same) keep their
historical "no pitch blob, latch once" path and emit byte-identical
ROM bytes.
The runtime gains two new pieces, both gated on a new
`__sfx_pitch_used` codegen marker so programs without varying-pitch
sfx pay zero bytes:
1. `gen_audio_tick` emits a per-frame pitch update block inside
the SFX tick: read a byte through `(AUDIO_SFX_PITCH_PTR),Y`,
write it to `$4002` (pulse-1 period low), advance the pointer.
The block bails on a zero high-byte pointer so a single
program can mix scalar-pitch and varying-pitch sfx without
one clobbering the other.
2. `emit_play_pulse` seeds `AUDIO_SFX_PITCH_PTR_LO/HI` with the
pitch-blob label for varying-pitch sfx and zeros it for
scalar-pitch sfx. The per-call branch is skipped entirely
when the program has no varying-pitch sfx anywhere.
The new `examples/sfx_pitch_envelope.ne` exercises the path with
a 16-frame siren sweep. Triangle and noise per-frame pitch are
deferred — they share the same data shape but the runtime ticks
for those channels still write only their volume registers, see
docs/future-work.md for the gap.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
Programs that put functions in switchable banks can now call across
bank boundaries — `bank A { fun step() { helper() } }` where
`helper` lives in `bank B` used to panic in the IR codegen. Three
small pieces unblock it:
1. **Generic trampoline.** `runtime/gen_bank_trampoline` no longer
takes a `fixed_bank_index` argument. Instead it reads the
caller's current bank from `ZP_BANK_CURRENT`, pushes it on the
hardware stack, switches to the target, JSRs the entry, then
pulls and restores the saved bank. The same per-callee stub
works for fixed→banked and banked→banked direction; nested
trampolines compose because each PHA/PLA pair sits inside its
own JSR/RTS frame. `gen_mapper_init` seeds `ZP_BANK_CURRENT`
with the fixed bank index for any banked mapper so the very
first cross-bank call from the fixed bank still restores to
the fixed bank (matching pre-banked-banked semantics).
2. **Codegen drops the panic.** The `Some(from), Some(to)` arm in
the call-resolution switch now emits `JSR __tramp_<name>` like
the fixed→banked case instead of panicking. Banked→fixed calls
still go direct (the fixed bank is always mapped at $C000).
3. **Bank-namespaced local labels.** Two banks emitting the same
`__ir_cmp_e_8` would trip the linker's discovery-pass duplicate-
label check the moment any banked code generated a comparison.
The new `local_label_suffix` helper prefixes the suffix with the
current bank name when banked code is being emitted, leaving
fixed-bank label generation untouched (so existing examples are
byte-identical apart from the trampoline / init bytes
themselves).
The new `examples/uxrom_banked_to_banked.ne` demonstrates the path
end-to-end: `bank Logic { fun step() { ... clamp() } }` calls
`bank Helpers { fun clamp() { ... } }` once per frame. The harness
golden is committed alongside it. The five existing banked example
ROMs change byte-for-byte because of the new trampoline shape and
the seed-ZP_BANK_CURRENT init, but their emulator goldens still
match exactly — observable behaviour is unchanged.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
Struct field types beyond the v1 scalar set (`u8`, `i8`, `u16`,
`bool`) used to error out with `E0201: struct fields must be
u8/i8/u16/bool`. The size accumulator already handled them
correctly — what was missing was: (1) the analyzer side that
synthesizes per-leaf symbols and allocations for nested structs
plus a single array-typed symbol for array fields, (2) the
parser's chained-field-access path, and (3) the IR-lowering
recursion through nested struct literal initializers and array
literal field values.
The synthetic-variable model carries through unchanged: a
`var p: Player` where `Player { pos: Vec2, hp: u8, inv: u8[4] }`
and `Vec2 { x: u8, y: u8 }` produces flat allocations for
`p.pos.x`, `p.pos.y`, `p.hp`, and `p.inv`, plus an intermediate
`p.pos` Struct symbol so dotted-name lookups still resolve. Array
fields get a single allocation with the array type so the
existing `Expr::ArrayIndex` lowering path handles `p.inv[i]`
without changes. Array-of-structs is still rejected with E0201
because the synthetic model can't index per-element layouts
without further codegen work.
The parser change is the only structural move: `parse_primary`
and `parse_assign_or_call` now loop the dot chain into a single
joined identifier so `p.pos.x` becomes `FieldAccess("p.pos", "x")`
and `p.inv[0]` becomes `ArrayIndex("p.inv", 0)`. The downstream
analyzer and IR lowering use the same `format!("{name}.{field}")`
join they already used for one-level access — no plumbing
changes required.
Includes a new `examples/nested_structs.ne` that exercises both
features end-to-end with two `Hero` instances carrying nested
positions and inventory arrays. The reproducibility tripwire
ROM is committed alongside it and the emulator harness has a
matching pair of golden files.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
The frame-overrun counter at $07FF was previously only readable
via `peek(0x07FF)`, which forces every program that wants to
guard against missed frames to know the magic address. This adds
two named query expressions:
- `debug.frame_overrun_count()` — cumulative miss count since reset
- `debug.frame_overran()` — sticky bit cleared by the next wait_frame,
so `debug.assert(not debug.frame_overran())` catches a miss in the
previous window without waiting for the counter to roll over.
The sticky bit lives at $07FE alongside the existing counter and
is set inside the same NMI-time overrun branch. Release builds
emit none of the runtime side: the NMI handler still skips both
writes, the codegen `wait_frame` only clears $07FE in debug mode,
and committed example ROMs stay byte-identical.
The new expression form parses through `parse_primary`'s `KwDebug`
arm, so the existing `debug.log(...)` / `debug.assert(...)`
*statement* parser stays untouched. The analyzer rejects unknown
methods with E0201 and stray arguments with E0203 so typos don't
silently compile to a zero load.
https://claude.ai/code/session_01KEczoNUX3WmcFLfq6iAQxB
The compile bench had a hand-maintained parallel copy of
`src/main.rs::compile`, and that copy went out of sync after
bank switching landed — the bench kept handing the linker
`PrgBank::empty(...)` slots even though the CLI started
populating per-bank instruction streams + trampoline requests.
The assembler then panicked with `unresolved label:
'__tramp_step_animation'` on `uxrom_user_banked.ne` under
`cargo test --all-targets`, which is what CI runs. A plain
`cargo test --release` (what CLAUDE.md used to document) never
builds the bench so the bug slipped through local validation.
Fix:
- New `nescript::pipeline` module with `compile_source(source,
source_dir, &CompileOptions)` that owns the full
`parse → analyze → lower → optimize → codegen → peephole →
link` pipeline including the per-bank stream + trampoline
reconstruction. Returns a `CompileOutput` carrying the ROM,
the linker result, analysis, IR, assets, instructions, and
source-loc markers so downstream tools have one place to
pull metadata from.
- `src/main.rs::compile` reduces to file I/O + preprocessing +
a single `compile_source` call + CLI-only side effects
(`--dump-ir`, `--call-graph`, `--asm-dump`, `--memory-map`,
`--symbols`, `--source-map`).
- `benches/compile.rs::compile_pipeline` becomes a one-line
`compile_source` call. It is now structurally impossible for
the bench to drift from the CLI path.
- `tests/integration_test.rs::compile_with_debug_artifacts`
likewise delegates to `compile_source`. This also fixes a
latent bug in the helper where it used `Linker::with_mapper`
without `.with_header(...)` — programs opting into
`header: nes2` would have quietly got an iNES 1.0 header
through this path.
- `CLAUDE.md`: updated the "Running the basics" section to
specify `cargo test --all-targets` (plain `cargo test` skips
benches) and to point at `scripts/pre-commit` with the exact
install command. Also installed the hook in this worktree.
All 24 existing `examples/*.nes` rebuild byte-identical through
the new pipeline. 624 tests + all 25 emulator goldens pass.
https://claude.ai/code/session_01MaNVcDmK9gsspRkdxowQAM
- Analyzer: new `W0108` warning when an array's byte size exceeds
256. The codegen lowers `arr[i]` to `LDA base,X` and the 6502's
X register is 8 bits, so elements past byte 255 are unreachable.
The old debug bounds check silently skipped arrays in that range;
it now clamps the compare to 255 and the analyzer diagnoses the
declaration up front.
- UxROM `__bank_select`: the routine previously wrote the bank
number to a fixed `$FFF0`, which works on emulators that don't
simulate bus conflicts (jsnes, Mesen permissive) but is broken
on real hardware because a single ROM byte can't match every
possible bank number. Fixed by `TAX; STA __bank_select_table,X`
— the store lands at `table + bank_num`, whose ROM byte is
exactly `bank_num`, so CPU bus = A = ROM = no conflict. New
`LabelAbsoluteX` addressing-mode variant in the assembler
resolves the table's base address through the existing fixup
pass. The two existing UxROM example ROMs shift a few bytes
but their goldens still match (jsnes is bus-conflict-permissive).
- Source maps: new `source_map_survives_aggressive_peephole_folding`
regression test. The reviewer was worried peephole could drop
`__src_<N>` labels and silently leave stale source-map entries.
Peephole actually treats labels as block boundaries and never
deletes them — the test pins that down by compiling a program
tailored to trip every peephole fold and asserting every
codegen-recorded source marker survives into the final linker
label table.
- Frame-overrun counter: new `debug_frame_overrun_counter_reads_back_from_user_code`
end-to-end test that proves the contract works: NMI emits
`INC $07FF`, user `peek(0x07FF)` lowers to `LDA $07FF`, and the
RAM allocator doesn't hand out `$07FF` to a user variable.
https://claude.ai/code/session_01MaNVcDmK9gsspRkdxowQAM
Writing $07 in `emit_play_triangle` and $0B in `emit_play_noise`
meant that a noise play following an in-progress triangle note
would clear bit 2 of $4015 and cut the triangle off mid-envelope
(and vice versa). Write $0F from both paths so every trigger keeps
pulse1, pulse2, triangle, and noise enabled; channels with no
active envelope stay silent via the runtime's per-channel counter
gating. Also fixes the attribute-byte packing comment in
`png_to_nametable` — the code was correct, the doc string had the
quadrant order backwards.
The only observable ROM change is `examples/noise_triangle_sfx.nes`
(two immediate operands shift) and its audio hash golden; the
committed PNG golden is byte-identical. Found in independent code
review after the section landed.
https://claude.ai/code/session_01MaNVcDmK9gsspRkdxowQAM
Adds a `bank Foo { fun bar() { ... } }` parser form so user functions
can opt into living in a switchable PRG bank instead of the fixed
bank, plus the IR codegen, runtime, and linker work to make calls
across the bank boundary actually run. Programs that don't use the
new syntax produce byte-identical ROMs to before — verified by
rebuilding every existing example and diffing.
Pipeline shape:
* Parser accepts both `bank Foo: prg` (legacy reserved slot) and
`bank Foo { fun ... }` (functions land in the named bank). Nested
functions get tagged `bank: Some("Foo")` on the FunDecl + IrFunction.
* Analyzer bumps the user zero-page start past `$10` whenever the
program declares any banked function, so `__bank_select`'s STA into
ZP_BANK_CURRENT can't clobber a user variable. Programs without
banked functions keep the legacy `$10` start.
* IrCodeGen emits each banked function into its own per-bank
instruction stream (`banked_streams: HashMap<String, Vec<Instruction>>`)
while the fixed-bank stream gets the dispatcher loop + state
handlers + top-level functions, exactly like before. Cross-bank
calls from the fixed bank rewrite `JSR __ir_fn_<name>` to
`JSR __tramp_<name>`; in-bank calls stay direct. Banked → fixed
calls are direct (the fixed bank is always mapped at $C000-$FFFF).
Banked → other-banked calls aren't supported in this pass and
panic loudly during codegen.
* Runtime's `gen_bank_trampoline` takes the trampoline label and
entry label as parameters now (one trampoline per banked function,
not one per bank) so the linker can request any number of stubs.
* Linker assembles banked banks twice: a discovery pass to learn
each bank's labels, then a final pass that seeds the merged label
table so banked code can JSR into the fixed bank's runtime helpers
(math, audio, etc.). The fixed-bank assembler is also seeded with
the cross-bank labels so the trampolines' `JSR __ir_fn_<name>`
resolves into the bank's $8000 window. New `asm::assemble_with_labels`
/ `asm::assemble_discover_labels` helpers wire this up.
* PrgBank carries `Vec<Instruction>` + a list of `BankTrampoline`
requests now, replacing the old `data: Vec<u8>` + single
`entry_label: Option<String>` shape. The compiler populates both
from the codegen output; the linker's two-pass assembly handles
the rest.
New example: `examples/uxrom_user_banked.ne` puts a sprite-stepping
helper inside `bank Extras { fun step_animation() { ... } }`. The
fixed-bank state handler calls it via the generated trampoline, and
the harness golden locks in pixel + audio output at frame 180.
UxROM is the only mapper exercised by the new example. MMC1 and
MMC3 also work through the same path (the linker emits the right
mapper-specific bank-select code), but no example uses them yet —
the existing `mmc1_banked.ne` / `mmc3_per_state_split.ne` keep
their fixed-bank-only layout.
Limitations carried forward:
* No banked → banked cross-bank calls (panics in codegen).
* No greedy size-packing; placement is explicit-only.
* MMC3 state handlers don't get banked (the per-state split path
is untouched).
