Replace the explored byte-layout-override engine (offset-ordered LLVM structs /
weld plans / byte-blobs — all unnecessary) with a much simpler design: a welded
`struct abi(.zig) extern compiler { … }` is a bodied header declaring its fields
in the bound compiler type's MEMORY order. The compiler reflects the real Zig
type (field names via @typeInfo, offsets via @offsetOf, size via @sizeOf —
nothing hand-maintained) and validates the header matches, with loud diagnostics.
On pass it is an ordinary struct whose natural layout already equals the Zig
layout — no reorder, no padding, no index/remap tables, no special LLVM path — so
@ptrCast'ing it to the compiler's own type and dereferencing is byte-identical.
When types.zig shifts, the header stops matching and the developer gets a specific
message to fix it.
- compiler_lib.zig: weldStruct reflects field names and bakes bound_types fields
in ascending-offset (memory) order; deleted computeWeldPlan/WeldPlan/WeldElement.
- nominal.zig validateWeldedStruct: precise diagnostics — field-not-found,
wrong-field-order (+ expected memory order), type-layout (size) mismatch,
total-size mismatch.
- Examples: 0627 (StructInfo in memory order, byte-identical, usable),
1186 (source-order StructInfo -> wrong-field-order diagnostic); 1183 refreshed.
- Design doc + checkpoint updated.
Introduce the welded comptime `compiler` library (`#library "compiler"` +
`abi(.zig) extern compiler`), per design/comptime-compiler-api.md, and unify
`callconv(...)` into the new `abi(...)` annotation.
abi(...) replaces callconv(...):
- New ABI enum { default, c, zig, pure }; `abi(.c|.zig|.pure)` parses in the
postfix slot before extern/export (and standalone). `kw_callconv` -> `kw_abi`.
- Migrated 52 sx files, the call-convention-mismatch diagnostic, and docs
(readme/specs) from `callconv(.c)` to `abi(.c)`.
Phase 1 — welded compiler library (parse -> registry -> validation -> bridge):
- `abi(.zig) extern compiler` parses on fn decls (carries abi/extern_lib) and
struct decls (StructDecl.abi/extern_lib).
- `#library "compiler"` is the comptime-only internal surface — never dlopen'd.
- src/ir/compiler_lib.zig: the binding registry (the safety boundary). `Field`
welded to StructInfo.Field with layout baked from the real Zig type
(@offsetOf/@sizeOf); `findType`/`findFn`. Welded structs are layout-validated
at registration (field set + total size) as a header checked against the impl.
- Host-call bridge: a `fn abi(.zig) extern compiler` dispatches under the
comptime interp to its registered Zig handler (intern/text_of round-trip),
never dlsym. IR Function.compiler_welded; validated in declareFunction.
- Comptime-only enforcement: a runtime call to a welded fn is a clean
build-gating error (emitCall), not an undefined-symbol link failure.
Phase 2.1 — byte-layout weld foundation:
- Decision: full byte-layout weld (sx struct laid out byte-identically to the
bound Zig type). Registered StructInfo (first non-natural / Zig-reordered
layout). `computeWeldPlan` — pure offset-ordered element plan + padding +
sx-field->LLVM-element remap; unit-tested. Emit/interp wiring is the next
sub-step (2.2+, see current/CHECKPOINT-COMPILER-API.md).
Examples: 0625/0626 (welded struct + fn round-trip), 1183/1184/1185
(layout-mismatch, unexported-fn, runtime-call diagnostics).
A generic ($T) -> Type type-fn comptime-evaluated only its return
EXPRESSION, so a local declared before the return ('vs := …; return
make_enum(…, vs)') was unresolved. Now a body with a prelude (statements
before the return) has its full body evaluated: createComptimeFunction-
WithPrelude lowers the pre-return statements into the comptime function's
scope before the return expr, so the locals resolve.
- comptime.zig: createComptimeFunctionWithPrelude (prelude stmts +
expr); evalComptimeTypeBody (extract prelude + return expr, scan the
whole body for declare() forward types); runComptimeTypeFunc factored
out of evalComptimeType (shared bail/declare-never-defined handling).
- generic.zig: route a type-fn body WITH a prelude through
evalComptimeTypeBody; no-prelude bodies stay on evalComptimeType (zero
change for RecvResult/TryResult etc.).
Non-generic builders (whole body already evaluated) and the List-growth
path are unaffected. Suite green (684).
TypeInfo gains a `tuple(TupleInfo) variant (TupleInfo{elements: []Type},
positional/unnamed) — completing the reflect/construct triad with enum
and struct.
- meta.sx: TupleInfo + `tuple TypeInfo variant.
- interp: reflectTypeInfo builds .tuple (tag 2) as bare type_tag elements
(no name pairs); defineType dispatches tag 2 -> defineTuple, which
decodes []Type and completes the declare slot as a structural .tuple
via replaceKeyedInfo (kind change). Tuples are structural so the
declared name is vestigial, but the slot is still completed in place so
define returns the handle (consistent with enum/struct).
- call.zig: the lower-time type_info guard now admits .tuple.
define(declare("P"), .tuple(.{elements=.[i64,f64]})) builds a tuple, and
define(declare("T"), type_info((i64,bool,f64))) round-trips one. Suite
green (683).
TypeInfo gains a `struct(StructInfo) variant (StructField{name,type});
the metatype system now reflects AND constructs structs, not just enums.
- meta.sx: StructField / StructInfo / `struct TypeInfo variant.
- interp: reflectTypeInfo builds .struct (tag 1) for a source @"struct";
define dispatches on the TypeInfo tag (defineType) -> defineEnum (0) /
defineStruct (1). defineStruct mirrors defineEnum (dup-field-name check
included) but completes the declare slot AS a struct via replaceKeyedInfo
(a kind change re-keys the intern map; updatePreservingKey asserts no
key change, true only for the enum path).
- call.zig: the lower-time type_info guard now admits @"struct".
define(declare("P"), .struct(.{ fields = .[ … ] })) builds a struct, and
define(declare("C"), type_info(SrcStruct)) round-trips one. Suite green
(682); enum path (0619) unchanged.
A bare declare("X") with no define left a zero-field nominal slot that
panicked at codegen (verifySizes: llvm_size != ir_size). evalComptimeType
now detects a zero-variant tagged_union result and emits a clean
build-gating diagnostic naming the type — a zero-variant enum is never a
legitimate construction result (defineEnum rejects empty variant lists
too). Self-reference (a declared slot completed by define) is unaffected.
Two same-named variants in a constructed enum silently succeeded —
construction (.a) and matching would ambiguously pick one. defineEnum
now bails when a variant name repeats, naming it. The name is dynamic so
it sets last_bail_detail directly (bailDetail takes a comptime string);
evalComptimeType renders it as a build-gating diagnostic.
`arr[lo..hi]` at comptime bailed for any non-string base — the interp's
.subslice op only handled string-backed values. Worse, the open-ended
`hi` came from a .length op that misread a 2-element array as a {ptr,len}
fat pointer (returning the 2nd element, not the count), so even lo/hi
weren't valid ints.
Fix, interp-only (runtime already handles arrays via LLVMTypeOf):
- Thread the base operand's IR type onto the Subslice op (base_ty); the
interp uses it to tell a bare array (elements = aggregate fields) from a
{data,len} slice (elements in the data field) — indistinguishable by
Value shape alone.
- Fold an open-ended slice's hi to the array's static length for fixed
arrays at lower time (runtime emitLength folds the same constant, so the
IR result is unchanged — no snapshot churn — but the comptime interp no
longer hits the ambiguous .length op).
- subsliceElements() resolves the element list (array/slice, inline or
slot_ptr-backed) and subslice returns a proper {data,len} slice value.
Suite green (678), no .ir changes.
evalComptimeType did `interp.call(...) catch return null`, dropping the
interp's last_bail_detail; callers poisoned to .unresolved with no
diagnostic, so the sentinel reached LLVM emission and panicked
("unresolved type reached LLVM emission"), or hid behind a downstream
cascade.
Clear last_bail_detail before the call; on the catch emit a build-gating
.err at the construction expr's span ("comptime type construction
failed: {detail}", mirroring the #run surfacing in emit_llvm.zig), then
return null to keep the .unresolved poison — now gated by a real message
so no unresolved type reaches emission unannounced.
Empty-variant define now prints 'comptime define(): enum has no
variants' and exits 1 (no panic); make_enum-style computed-slice
failures show their root reason at the construction site.
type_info reflects an enum / tagged-union INTO a TypeInfo value — the
inverse of define's decode — so define(declare(n), type_info(T)) mints
a byte-identical copy with NO literal variant list.
- inst.zig: new BuiltinId.type_info (comptime-only, like declare/define).
- lower/call.zig: replace the 'not yet implemented' bail. Resolve $T at
lower time, reject non-enum/non-tagged-union loudly with a good span,
emit callBuiltin(.type_info, [const_type], TypeInfo).
- interp.zig: reflectTypeInfo builds the exact nested-aggregate Value
defineEnum decodes — variant {name,payload}, slice {data,len}, EnumInfo
{variants}, TypeInfo {tag0, EnumInfo}. tagged_union reflects field.ty
(tagless already void); payloadless `enum` reflects void per variant.
- emit: unchanged — type_info is always comptime-evaluated, the existing
comptime-only else arm (shared with declare/define) never fires.
0619 turns green: a source enum (circle:f64 / rect:i64 / empty) reflected
and reconstructed, constructs and matches like the original.
A nominal aggregate that contains itself (or a mutual peer) BY VALUE has no
finite layout and infinite-recursed typeSizeBytes into a stack overflow —
for SOURCE enums/structs as well as comptime-constructed types.
New `checkInfiniteSize` pass (lower/decl.zig, Pass 1g — after type
registration, before body lowering): walks the by-VALUE containment graph
(pointer/slice/optional payloads break the cycle, so `*Self` stays valid);
on a back-edge it emits a loud diagnostic — "type 'X' is infinitely sized
(it contains itself by value); use a pointer ('*X') to break the cycle" —
and poisons the offending field to `.unresolved` so sizing can't recurse
before the build halts on the error. Covers source + declare/define types,
direct + mutual recursion.
examples/1178 locks the diagnostic; issue 0139 marked RESOLVED. This also
completes METATYPE PLAN F5's by-value-self-reference rejection. Full suite
green (675).
declare now takes the type's NAME — `declare(name) -> Type` — because the
compiler needs it at compile time to register the forward type, which is
what makes self-reference resolve. EnumInfo drops `name` (it lives on
declare now); define completes the handle's body in place (the slot is
already named).
Self-reference mechanism (evalComptimeType): before lowering a comptime
type expression, preregisterForwardTypes scans it (and a called ctor fn's
body) for `declare("Name")` calls and registers each as an empty forward
nominal type AND binds it as a type alias. The alias is essential: a
`Name :: ctor()` decl makes `Name` a const_decl author, so a `*Name`
self-reference resolves through the forward-ALIAS path
(type_aliases_by_source), which a bare findByName registration doesn't
satisfy. With both in place `*Name` resolves to the forward slot at lower
time; the interp's declare returns that same slot; define fills it.
List :: make_list();
make_list :: () -> Type {
h := declare("List");
return define(h, .enum(.{ variants = .[
EnumVariant.{ name = "cons", payload = *List },
EnumVariant.{ name = "nil", payload = void } ] }));
}
Verified: cons/nil construct + match (direct and through the pointer),
multi-node list traversal via a recursive `count(*List)`. meta.sx
RecvResult/TryResult + examples 0614/0615/0617 updated to declare(name);
full suite green (673).
Per the directive to strip reify entirely: the sx `reify(info)` one-shot is
removed. `define(handle, info)` now RETURNS the (completed) handle, so the
one-shot constructor chains as a single expression:
T :: define(declare(), .enum(.{ name = "T", variants = ... }));
- meta.sx: drop reify; RecvResult/TryResult use `define(declare(), …)`.
- interp .define returns the handle type_tag (was void); call.zig lowers it
with `Type` result and sets the info arg's target type to TypeInfo so the
intercepted call still infers the `.enum(…)` literal.
- returnExprMintsType: a type-fn body that returns `define(…)` (or a bodied
non-generic Type-returning sx helper) is comptime-evaluated.
- examples 0614 (direct) + 0615 (type-fn) use `define(declare(), …)`.
Full suite green (673). Files/docs still carry the old reify naming — the
rename sweep is the next commit.
Second slice of the re-architecture — the compiler now has ZERO type-
construction code beyond declare/define.
- instantiateTypeFunction: a type-fn body returning a computed Type (a call
to a non-generic, bodied, Type-returning fn) is comptime-evaluated with the
type bindings active, then renamed to the mangled instantiation name for
identity (renameNominalType). Replaces the old reify-call pattern-matching.
- DELETED: reifyType (lower/nominal.zig), findReturnReifyCall (lower/generic.zig),
and the stale inline-position reify gate in resolveTypeCallWithBindings.
- evalComptimeType (was evalComptimeTypeNamed): pure eval, no rename; the
type-fn caller renames explicitly. renameReifiedType → renameNominalType.
- The TYPE NAME now travels in the data: EnumInfo gains `name`, and define()
names the slot from it (the compiler derives no name from a binding LHS).
examples/0614/0615 carry `name = "..."`; RecvResult/TryResult set it too.
- field_type stays a reflection #builtin (reads a type); only construction
moved out. All reify mentions stripped from compiler source.
examples 0614/0615/0617 run on the floor. Full suite green (673).
First slice of the re-architecture. The compiler gains two comptime
type-construction builtins — declare() (mint an empty/undefined nominal
slot) and define(handle, info) (decode a TypeInfo VALUE + complete the
slot) — executed by the interpreter against a new `mint` TypeTable handle
(setMintTable). reify becomes PLAIN sx in meta.sx:
reify :: (info) -> Type { h := declare(); define(h, info); return h; }
`E :: f(...)` where f is a non-generic Type-returning fn (reify, and later
make_enum) is now comptime-evaluated via evalComptimeTypeNamed: wrap the
call in a throwaway comptime fn, run it through the interp with the mint
table enabled so declare/define mint the type, read back the type_tag, and
rename the anonymous slot to the binding name. The compiler has ZERO reify
knowledge at the decl site — the old `E :: reify` hook is deleted.
examples/0614 (inline reify) now runs on this floor. Full suite green (673).
INTERMEDIATE: reifyType + findReturnReifyCall still serve the type-fn path
(0615/0617) and will be deleted in the next slice (type-fn body
comptime-eval), after which the compiler has no reify code at all.
REIFY Phase 2.1. fieldTypeOf (lower/generic.zig, re-exported on Lowering)
returns the i-th member type of T: struct field / tagged-union + union
variant payload (.void for a tagless variant) / tuple element / array +
vector element. Out-of-range and memberless types poison to .unresolved
with a loud diagnostic (never a silent default). Wired into
resolveTypeCallWithBindings (replacing the Phase-2 bail); since it folds
to a TypeId at lower time it composes inside type_eq / type_name / any
type-arg slot.
examples/0616 green: struct fields (name via field_name + type via
field_type), type_eq fold, tagged-union payloads incl. quit -> void.
Suite green (672 examples, 447 unit).
type_info($T) -> TypeInfo (reflect into a value, inverse of reify) is
NOT done — still bails loudly; it's the larger Phase 2.2 step (widen the
TypeInfo data model + comptime value construction). Plan/checkpoint updated.
REIFY Phase 1.1 (Phase 1 complete). instantiateTypeFunction detects a
type-fn body that returns reify(...) (findReturnReifyCall) and routes it
to reifyType under the instantiation's name — mangled for inline use,
the alias name for `Foo :: Box(i64)` — with the type-arg bindings active
so reify payloads (`payload = T`) resolve against the instantiation args.
Placed before the general case, whose resolveTypeWithBindings would
route the reify call to the inline-position loud bail.
Registering under the mangled name lets the top-of-instantiation cache
return the SAME TypeId on a second instantiation, so Box(i64) resolved
at two independent sites is ONE type (Contract 1). examples/0615 green
(build()->consume() cross-site + `b : Box(i64) = .none`). Suite green
(671 examples, 447 unit).
REIFY Phase 0.2 (Phase 0 complete). Lowering.reifyType (lower/nominal.zig)
reads the flat-enum TypeInfo literal off the AST, synthesizes an
ast.EnumDecl, and feeds it through the SAME type_bridge.buildEnumInfo
path source enums use — so the minted type is byte-identical to a
hand-written `enum { value: i64; closed; }` and flows through enum
codegen (layout / construct / match) UNMODIFIED (Contract 2).
Wired at the `E :: reify(...)` const-decl hook in lower/decl.zig
(replacing the Phase-0.0 loud bail). Unsupported argument shapes bail
loudly via reifyBail — never a silent default. The generic.zig inline
reify path now reports it's only supported in a `::` binding (Phase 0).
examples/0614 green: reify a {value: i64, closed} enum, construct
.value(3) and .closed, match both -> "value 3" / "closed". Full suite
green (670 examples, 447 unit).
REIFY Phase 0.0. Add the comptime type-metaprogramming surface as the
on-demand module modules/std/meta.sx (NOT the prelude — declaring its
data types in always-loaded core.sx interns them into every module's
type table and shifts every .ir snapshot):
- EnumVariant / EnumInfo / TypeInfo data types. TypeInfo's variant uses
the backtick raw escape `enum so it reads as the keyword.
- reify / type_info / field_type as bodyless #builtin decls.
Each builtin bails LOUDLY when reached unimplemented (no silent default):
- reify(...) in a :: type-alias position -> decl.zig .call branch
(also the Phase 0.2 construction hook); poisons the alias .unresolved.
- reify / field_type in any other type position ->
generic.zig resolveTypeCallWithBindings.
- type_info(...) in expression position -> call.zig tryLowerReflectionCall.
Unit test src/parser.test.zig (registered in root.zig) locks that the
decls parse. zig build test green (447 unit, 669 examples).
Drive a bundled `zig` as `zig cc` for the AOT link step, supplying lld + CRT
+ libc (musl/glibc/mingw) so `sx build` produces native binaries with no host
toolchain. Default Linux output is static musl (portable-anywhere).
- src/zig_backend.zig: discover zig ($SX_ZIG / bundled-next-to-exe / PATH);
bundled-vs-PATH provenance gates auto-activation.
- src/target.zig: selectZigLinker + emitZigLinkArgv + zigTargetTriple, dispatched
before the per-OS branches; macOS/Linux/Windows in scope.
- src/ir/emit_llvm.zig: LLVMNormalizeTargetTriple so vendor-less zig triples
(e.g. x86_64-windows-gnu) parse to the correct OS/object format (COFF not ELF).
- src/main.zig: --self-contained / --no-self-contained; linux-musl, linux-musl-arm,
windows-gnu shorthands; de-vendor linux/linux-arm to match the corpus runner.
- examples/1660: Windows Win32 print-42 + exit(0) via kernel32 (ir-only off-Windows).
Auto-activates only for a bundled zig; a PATH-only zig engages under
--self-contained, so native dev/CI builds are never silently rerouted.
Docs: readme Cross-Compilation, design/bundled-zig-link-backend-design.md, current/PLAN-DIST.md.
A `"s"` input operand feeds a function/global symbol; the template's
%[name] emits the platform-mangled name, so `bl %[fn]` / `call %[fn]`
branches DIRECTLY to it (PC-relative, no register load — one fewer
indirection than register-indirect `blr`).
Lowering: an `"s"` input lowers its RHS normally (a function name →
`ptr @fn`); the rejection added last commit is removed. Emit: a symbol
operand is passed with its OWN llvm type (LLVMTypeOf) and no coercion —
the function value is a `ptr`, and the old coerce-to-register-int path
mistyped it and failed the verifier. New asmIsSymbol helper.
Verified on aarch64: examples/1656 (sx → asm → bl _cb → sx → 42); the
emitted asm is a direct `bl <_cb>` (objdump-confirmed), IR constraint
`...,s,...`(ptr @cb). Flipped 1656 from the rejection lock to a runnable
aarch64 example. zig build test green (665 corpus, 446 unit).
A symbol operand (constraint "s") feeds a function/global symbol whose
mangled name the template emits — enabling a DIRECT `bl %[fn]` (one
fewer indirection than register-indirect `blr`). Until now `"s" = fn`
fell through to emit and produced an LLVM-verifier crash (param type
mismatch). Reject it at lowering with a clear diagnostic instead, and
lock that with examples/1656-platform-asm-symbol-operand.sx. The next
commit implements it and flips the example to run (→ 42).
Implements indirect-memory (`=*m`) `-> @place` outputs — the last
substantive asm feature. Unlike a write-through `=` output (which
returns a value that is then stored), an indirect output passes the
place ADDRESS to the asm and the asm writes through it; there is no
return slot.
emitInlineAsm:
- indirect outputs are excluded from the LLVM return type;
- their pointer is passed as an opaque `ptr` call arg, placed FIRST
(the arg-consuming constraint order is: output-section indirect
pointers, then inputs, then read-write tied seeds);
- each indirect arg gets an `elementtype(T)` call-site attribute
(required in the opaque-pointer era), T = the pointee type;
- the store-back loop skips indirect outputs (already written).
New asmIsIndirect helper. Lowering stops rejecting `*` (constraint kept
verbatim; `=*m` reaches the constraint string as-is). asmOperandIndex
is unchanged — indirect outputs still count as operands, so `%[name]`
${N} numbering holds.
Verified by running on aarch64: store-through-pointer (str x9, %[out]
→ 42, IR `=*m,~{x9}` with `ptr elementtype(i64)`) and a mixed case
(indirect + value output + input → `=*m,=r,r`, indirect ptr arg first,
${0}/${1}/${2} correct). 1652 flipped from the rejection lock to a
runnable aarch64 example (ir-only elsewhere). zig build test green
(661 corpus, 446 unit).
A scalar `::` constant folds to its value and has no storage. The
unary `.address_of` lowering (src/ir/lower/expr.zig) skipped the
alloca path (is_alloca == false) and resolveGlobalRef (scalar consts
get no storage global), falling through to the generic addr_of arm,
which reinterpreted the folded value as a pointer:
`inttoptr (i64 <value> to ptr)`. That wild pointer segfaulted on
deref and emitted invalid stores for inline-asm `-> @const`.
Diagnose instead, in the address_of(identifier) path: a non-alloca,
non-ref-capture, non-pack-elem scope binding (local scalar const) and
a module_const_map name not backed by storage (module scalar const)
both report "cannot take the address of constant '<name>' — a scalar
'::' constant has no storage …" and return a placeholder Ref. Chose
diagnose over materializing read-only storage (consistent with the
fold-only scalar model). Array/struct consts keep real storage and
stay addressable (@K/@LIT unchanged).
Also gives the ASM stream's planned output-to-const rejection for
free — asm `-> @const` lowers through the same path. Regression:
examples/1177-diagnostics-addr-of-const-rejected.sx. Resolves 0138.
Implements read-write (`+r` / `+{reg}`) `-> @place` outputs. LLVM has
no `+` constraint, so a read-write place lowers to:
- an output `=` constraint (return slot, stored back through the
place after the call), with the leading `+` rewritten to `=`; plus
- a TIED input constraint (the decimal index of that output) appended
after the regular inputs, seeded with the place's loaded value
passed as a call arg.
Tied inputs are appended last so existing operand indices (%[name] ->
${N}) are undisturbed; asmOperandIndex stays correct. Lowering no longer
rejects `+` (indirect `*` still rejected). emitInlineAsm grows the
arg/param arrays by the rw count, loads each seed, and emits the tied
constraint.
Verified by running: increment-in-place (41 -> 42) and a mixed case
(rw place + regular input + value output) producing the textbook
"=r,=r,r,0" constraint with correct ${N} indices. 1650 flipped from
the rejection lock to a runnable aarch64-pinned example (ir-only
elsewhere). zig build test green (658 corpus, 446 unit).
An asm result can be STORED through a place (a local / struct field) instead of
returned; the place output does not join the result tuple.
- parser.zig: `-> @place` parses `@place` as an ordinary address-of expression
→ an out_place operand (the in-function form; reuses the existing `@` prefix).
- inst.zig: AsmOperand gains out_ty (the output slot's value type) so emit can
build the combined return struct without re-deriving from Inst.ty.
- lower/expr.zig: out_place operand = the lowered @place address, out_ty = the
pointee. Read-write (`+`) and indirect-memory (`*`) constraints rejected loudly
(not yet implemented) rather than miscompiled.
- ops.zig emitInlineAsm: the LLVM return type is built from ALL outputs
(out_value + out_place); after the call, out_place slots are stored through
their address and out_value slots rebuild the sx result. Fast path when there
are no place outputs (the struct return IS the result — pure-value asm IR
unchanged).
Verified: write-to-local (42), struct field, mixed value+place (v=10 b=20), `+`
rejected. Locked with 1649-platform-asm-place-output (mixed, runs on aarch64).
zig build test green (657 corpus, 446 unit).
A top-level `asm { "tmpl", };` block (template only) lowers to LLVM `module asm`;
a lib-less `extern` declaration calls into the symbols it defines (the import
direction reuses the existing C-FFI extern path — no new surface).
- ast.zig: asm_global node (AsmGlobal { template }).
- parser.zig: parseAsmGlobal, dispatched from parseTopLevel on kw_asm — rejects
`volatile` and any operands/clobbers (template only). The in-function asm
expression form stays in parsePrimary.
- module.zig: Module.global_asm list; lower/decl.zig captures each template in
lowerMainAndComptime (the real top-level pass — lowerDecls is dead for
top-level); emit_llvm.zig emit() appends each via LLVMAppendModuleInlineAsm in
source order.
- the new node forced asm_global arms in sema.zig (analyzeNode +
findNodeAtOffset) and semantic_diagnostics.zig (checkBindingNames).
Verified end-to-end: an aarch64 `_my_add` global routine, called via `extern`,
returns 42 — AOT only (the ORC JIT doesn't link module-asm symbols; global-asm
symbols live in the final linked binary). Locked with 1648-platform-asm-global
({ "aot": true, "target": "macos" } → AOT build+run on aarch64, ir-only else).
zig build test green (656 corpus, 446 unit).
Replaces the N>1 "Phase E" bail with a shared asmResultType helper (lowering +
inferType) that derives the result type from the out_value operands: 0→void,
1→T, N→a named tuple (fields named via the §II.5 effective-name rule).
Key realization: toLLVMType(tuple) already produces a literal struct {T1,…,Tn} —
exactly what LLVM's multi-output inline asm returns — so emit needs NO change.
Building the op with a tuple result type makes the asm call return the struct,
which IS sx's tuple value (destructured by the normal tuple_get path).
inferType's .asm_expr arm now also delegates to asmResultType (single owner), so
`return asm`, `x := asm`, and `q, r := asm` all agree on the type.
Verified end-to-end on aarch64: split(0x1234)→(lo=52,hi=18), a udiv/msub
divmod→(3,2). IR: `call { i64, i64 } asm "divq ${4}",
"={rax},={rdx},{rax},{rdx},r,~{cc}"(…)` → extractvalue → tuple.
1640 → the x86_64 multi-output IR lock (ir-only); 1647 → a multi-output example
that runs on aarch64.
zig build test green (655 corpus, 446 unit).
lowerAsmExpr stops bailing and builds the inline_asm op: resolves each operand's
effective name (§II.5 — explicit [name] else the {reg} pin), interns
template/constraints/clobbers, lowers input Refs, derives the result TypeId
(0→void, 1→T). Adds the last deferred validation (every %[name] must name an
operand). Multi-output (N>1) bails with a named "Phase E" diagnostic.
emitInlineAsm (backend/llvm/ops.zig) ports Zig's airAssembly: assembles the LLVM
constraint string (outputs → inputs → ~{clobber}, ',' → '|'), rewrites the
template (%[name]→${N}, %%→%, $→$$, %=→${:uid}), then LLVMGetInlineAsm +
LLVMBuildCall2 (AT&T dialect). Dispatch wired in emit_llvm.zig (replacing the C.0
@panic tripwire).
inferType gains an .asm_expr arm (expr_typer.zig) so a bare `x := asm {…-> T}`
binding types correctly — without it the binding inferred .unresolved and
silently produced 0.
llvm_shim.c: LLVMInitializeNativeAsmParser() — the JIT must assemble inline asm
at run time.
Verified end-to-end on the aarch64 host: `mov`/`add` with register-class inputs
and a value output run (exit 42/99), `nop volatile` runs (exit 0). IR is
textbook: `call i64 asm "add ${0},${1},${2}", "=r,r,r"(…)`.
Locked with 1645 (aarch64 add, runs; ir-only on non-aarch64) + 1646 (:= binding).
Updated 1640 (now Phase-E bail) + 1642 (now runs).
zig build test green (654 corpus, 446 unit).
Adds the `inline_asm: InlineAsm` opcode to the IR Op union (inst.zig): interned
template + operand list (role/name/constraint/operand) + interned clobber names
+ has_side_effects; the result rides on Inst.ty (void / scalar / tuple).
The new variant forces coverage in the exhaustive Op switches:
- interp.zig: loud bailDetail — inline asm is never comptime-evaluable.
- print.zig: an IR-dump arm.
- emit_llvm.zig: a @panic TRIPWIRE — emit lands in Phase D, and until then
lowerAsmExpr still bails, so no inline_asm op is ever created. Reaching emit
would mean lowering switched over before emit was ready; crash loudly rather
than miscompile.
No behavior change: lowering still bails, the op is constructed only in the new
`inline_asm op shape` unit test (inst.test.zig).
zig build test green (652 corpus, 446 unit).
Extends lowerAsmExpr with a pinnedRegister(constraint) helper and two §II.5
operand-naming checks, in the compile path before the codegen bail:
- reject the echo form `[eax] "={eax}"` — a label identical to the register its
own constraint pins is redundant (the operand is already auto-named after the
register); the useful form is a label that differs (`[quot] "={rax}"`);
- reject duplicate operand names (ambiguous %[name] / result field).
Locked with 1643-platform-asm-echo-name and 1644-platform-asm-duplicate-name.
zig build test green (652 corpus, 445 unit).
Restructures the .asm_expr lowering arm into lowerAsmExpr, which validates the
asm shape with specific named diagnostics BEFORE the not-yet-implemented codegen
bail, so the user sees the real problem first. Two checklist items enforced:
- template must be a compile-time-known string ("..." or #string), not a
runtime expression;
- an asm with no value outputs must be `volatile` (else its effects could be
deleted) — mirrors Zig's rule.
Valid shapes still bail with the "codegen not yet implemented" message. Result-
type derivation + the operand auto-naming rule stay deferred to Phase C, where a
real IR op makes the result type observable/testable.
Locked with 1641-platform-asm-missing-volatile (the volatile error) and
1642-platform-asm-nop-volatile (no-output + volatile accepted → codegen bail).
zig build test green (650 corpus, 445 unit).
`asm volatile? { "tmpl", [name]? "constraint" (-> Type | = expr), …,
clobbers(.…) }` now parses into a flat-operand AsmExpr/AsmOperand (ast.zig +
parser.zig parseAsmExpr, dispatched from parsePrimary on .kw_asm). `volatile`
and `clobbers` are recognized contextually (not reserved). `-> @place`
write-through is rejected with a clear "Phase 2" parse error.
Codegen is not implemented yet (IR op + LLVM emit are Phases C–E), so lowering
bails LOUD + named via an explicit .asm_expr arm in lower/expr.zig (not the
generic unknown_expr else) — emitPlaceholder makes hasErrors() abort the build
on the message.
The new asm_expr tag forced (and got) arms in three exhaustive Node.Data
switches: sema.zig analyzeNode + findNodeAtOffset, semantic_diagnostics.zig
checkBindingNames — each recurses into template + operand payloads.
Design: adopted the operand auto-naming rule (design §II.5) — name auto-derived
from a {reg} pin, explicit [name] only when it differs or for register-class
operands, echo form rejected. Typing-stage rule; parser stores name: ?[]const u8.
Locked with examples/1640-platform-asm-parse.sx (multi-output divmod: named
operands, register pins, clobbers — parses then bails, called from main).
Also files issue 0137 (pre-existing, orthogonal: `sx run` with no `main`
segfaults via an unguarded JIT entry lookup in target.zig — not an asm bug).
zig build test green (648 corpus, 445 unit).
Reword every 'foreign' comment to the extern/runtime-class vocabulary matching the
renamed identifiers (foreign call→extern call, foreign class→runtime class, foreign
path→runtime path, the #foreign-literal comment mentions → extern, etc.). Also fixes
two USER-FACING issues: the 'expected … #foreign … after type annotation' parse error
no longer advertises the removed keyword, and the Android 'no #jni_main' help
diagnostic now shows '#jni_class(…) extern' instead of the rejected '#foreign
#jni_class'. Removed the now-dead prefix-#foreign-vs-postfix conflict branch in
parseRuntimeClassDecl (the caller rejects #foreign before it runs).
src/ now contains 'foreign' ONLY in the hash_foreign token machinery + its 4
rejection messages — the deprecation mechanism (kept per the 9.0 recommendation; the
message MUST name #foreign to guide migration). Snapshot-neutral; suite green
(646 corpus / 444 unit, 0 failed).
The JNI/runtime-class path (Decision 5, Runtime* family). Coordinated across the
hook boundary so the BuildOptions accessor + its registered hook string stay in sync:
- src/: RuntimeClassDecl.foreign_path→runtime_path, splitForeignPath→splitRuntimePath,
foreignPathToJavaName→runtimePathToJavaName, jni_main_foreign_paths→
jni_main_runtime_paths, hookJniMainForeignPathAt→hookJniMainRuntimePathAt, and the
hook string 'BuildOptions.jni_main_foreign_path_at'→'…runtime_path_at'.
- library/: build.sx accessor jni_main_foreign_path_at→jni_main_runtime_path_at +
bundle.sx call sites + the local var → runtime_path + a comment.
- specs.md: the accessor name + <foreign_path_with_dots> doc refs.
- Regenerated 37 .ir snapshots: every program importing build declares the renamed
@BuildOptions.jni_main_runtime_path_at hook stub — symbol-name change only (verified
the .ir diff is ONLY this rename; reverted orthogonal empty-file normalization).
Suite green (646 corpus / 444 unit, 0 failed).
Per user feedback: don't introduce new terminology. The RuntimeClassDecl
reference-vs-define flag (set by the postfix 'extern' modifier, == old prefix
'#foreign #objc_class') is named is_extern, matching the keyword that drives it
and the existing is_extern on VarDecl/IR. Renamed is_reference→is_extern,
is_reference_eff→is_extern_eff; updated the field comment. Snapshot-neutral; green.
The runtime-class object-model identifiers (Decision 5): parse/lower/find/resolve/
register/stamp fns Foreign→Runtime (parseRuntimeClassDecl, lowerRuntimeMethodCall,
findRuntimeMethodInChain, resolveRuntimeMethodReturnType, registerRuntimeClassDecl,
runtimeClassStructType, runtimeKindForOffset, …); state foreign_class_map→
runtime_class_map, current_foreign_class/_method→current_runtime_*, the
foreign_class_decl union variant→runtime_class_decl, foreign_method/static/instance/
class→runtime_*; and the reference-vs-define flag is_foreign→is_reference (+
is_foreign_eff→is_reference_eff) now that it only lives on RuntimeClassDecl.
Snapshot-neutral; suite green (646/444).
Remaining 9.2: the foreign_path family (coupled .sx hooks: jni_main_foreign_path_at
spans build.sx/bundle.sx/compiler_hooks.zig/specs.md) + the extern-ref validators
(checkForeignRefs etc. → Extern, linkage not runtime) + bare 'foreign' comments.
The last linkage-family 'foreign' carrier. Migrated c_import.zig auto-synthesis
(#import c {#include}) to build the extern shape (empty-block body + extern_export
= .extern_) instead of a foreign_expr body — the Phase 5.0 fn-body flip applied to
auto-synth. With nothing left building it, deleted the foreign_expr union variant +
ForeignExpr struct (ast.zig) and every reader: the dead-arm switch cases (sema,
resolver, generic, call, semantic_diagnostics, lsp), the coalescing reads in
decl.zig (is_foreign local, cc/rename/dedup/variadic/visibility gates) + pack.zig,
and checkForeignRefs (now reads extern_lib only). 9.1 LINKAGE PURGE COMPLETE — all
that remains in src/ is the runtime-class family (9.2) + comments. Snapshot-neutral
(the #import c examples 1215/1216/1217 + sqlite 1624 exercise the synth path); suite
green (646 corpus / 444 unit, 0 failed).
VarDecl carried BOTH the legacy is_foreign/foreign_lib/foreign_name AND the new
is_extern/extern_lib/extern_name (parallel forms coalesced during the migration).
The global #foreign parse path now rejects, so the legacy trio is write-dead and
read in only 3 coalescing sites (decl.zig). Simplified those readers
(vd.extern_name orelse vd.name; vd.is_extern) and deleted the dead fields. Build
confirms no other setter/reader. Snapshot-neutral; suite green (646/444).
Remaining linkage (9.1): foreign_expr (25, still built by c_import.zig auto-synth)
+ ForeignClassDecl.is_foreign (runtime-class, → 9.2). Runtime-class family (9.2,
Decision 5) is the big remaining src/ rename.
The dup-C-symbol diagnostic (decl.zig) and the resolveFuncByName panic (call.zig)
now say 'extern symbol' instead of 'foreign symbol' — the keyword-neutral internal
wording catches up to the extern-only surface. Intentional snapshot regen of 1172
(the only assertion of this message). Suite green (646/444).
Mechanical src/ rename of the linkage-family identifiers whose extern_* target is
collision-free: callForeign→callExtern, marshalForeignArg→marshalExternArg,
dedupeForeignSymbol→dedupeExternSymbol, foreign_name_map→extern_name_map,
is_foreign_c_api→is_extern_c_api. Snapshot-neutral (internal only); suite green
(646 corpus / 444 unit, 0 failed).
Deferred (need per-site analysis — target name already exists): is_foreign↔is_extern
(38 existing), foreign_lib/foreign_name↔extern_lib/extern_name (15/16 existing),
foreign_expr (still built by c_import.zig auto-synthesis). Runtime-class family
(ForeignClassDecl etc. → Runtime*, Decision 5) is Phase 9.2.
The prefix #foreign linkage directive is removed. All four parse sites
(const-with-type, data global, fn body, runtime-class prefix) now reject it with
a migration message ('#foreign has been removed; use the postfix extern (import) /
export (define) linkage keyword instead'); added a span-aware failAt for the
runtime-class case (the lookahead consumes the token before the reject decision).
Greens the Phase 8.0 xfail 1176.
- Deleted obsolete tests: 1174 (#foreign+postfix conflict — unreachable now that
#foreign alone is rejected) and 1620 (#foreign nosuchunit lib-ref — superseded by
the extern twin 1231). Their assertions tested #foreign-specific behavior.
- Removed the GATE A→B unit test + lowerSrcToIr helper (lower.test.zig): it locked
#foreign ≡ extern through the migration; with #foreign gone there is nothing to
compare. Converted the in-source 'parse void function with foreign body' parser
test to the surviving postfix 'extern' spelling (identical resulting AST).
- specs.md + readme.md drop #foreign; document extern/export as the sole C-linkage
surface.
extern_export in parseFnDecl is now const (the fn-body arm that mutated it is gone).
Suite green (646 corpus / 444 unit, 0 failed). NOTE: comment-only #foreign in
examples + issues/*.md prose + internal foreign_* identifiers remain for Phase 9
(now unblocked: Decision 6 = purge everything).
Phase 5.0 flipped the fn-decl and data-global #foreign parser paths onto the
same extern-named AST that postfix extern produces, so the A→B gate's fn/global
cases are now STRUCTURALLY identical (guaranteed by construction, not empirically
equal). Annotate the gate header to record this and keep it as a regression
tripwire against a future reader re-diverging the two spellings or a revert of
the flip. Add a fn-rename case (extern_name axis: c_abs -> "abs") to broaden
coverage beyond bare import. Test-only; suite green (647 corpus / 444 unit, 0
failed). PHASE 5.1 COMPLETE → PART B Phase 5 done; next Phase 6 (migrate stdlib).
isPlainFreeFn / isPlainFreeFnDecl excluded a #foreign body but classified
an empty-block extern fn as a plain free function, so existing extern fns
were wrongly counted in the bare-call ambiguity verdict (and eligible for
the out-of-line-slot / shadow-author pass). Both predicates now also
exclude extern_export == .extern_ (an external C symbol with no
sx-lowerable body, name-keyed first-wins dispatch like #foreign); export
keeps a real body and stays plain-free. Greens example 1230 — same-name
extern authors compile like their #foreign twins (0729).
646 corpus / 444 unit, 0 failed.
Two gates were keyed on the `#foreign` (foreign_expr) body shape only:
- declareFunction: the is_variadic drop (decl.zig) — a variadic extern
kept its trailing slice param in the IR signature.
- packVariadicCallArgs: the call-site early-out (pack.zig) — extras were
slice-packed instead of passed through the C `...` slot.
Both now also fire for `extern_export == .extern_`, so a variadic
`extern` drops the trailing `..args: []T`, sets is_variadic, and passes
extras through the C ABI with default argument promotion — byte-identical
to its `#foreign` twin. Greens example 1229.
645 corpus / 444 unit, 0 failed.
The non-transitive C-import visibility gate (`isVisible(.c_import_bare)`)
only recognised the legacy `#foreign` body shape; a bare `extern` fn
(empty-block body + extern_export == .extern_) escaped via the
`body != foreign_expr -> return true` arm and was caught only by the
general isNameVisible gate, yielding the generic 'not visible' wording
instead of the C-specific 'C function not visible; add #import' one.
Now both lib-less spellings route to visibleOverEdges, and a library-
bound `extern LIB` (like a library-bound `#foreign LIB`) stays
unconditionally visible. This makes a future fn-decl `#foreign`->`extern`
migration byte-identical at this gate. Greens example 1228.
644 corpus / 444 unit, 0 failed.
The define path now honors the optional `export … "csym"` symbol-name
override (gap iii). declareFunction's rename branch fires for `export` too:
the extern stub is declared under the C name and the sx→C mapping recorded
in foreign_name_map. lazyLowerFunction then resolves the stub by that C
name (via foreign_name_map) so the body promotes into the C-named function
— emitting `define @triple_c` instead of `@sx_triple`. sx-side call sites
to the sx name resolve through the same map (verified: 5*5 prints 25).
example/1227 greens: the companion C calls `triple_c` and prints
call_triple(7) = 22. Bare export (1226) is unaffected (no rename → sx
name). Suite green (638 corpus / 443 unit). Phase 2 (`export`) complete.
