Surfaced while writing the Phase 1.11 in-construct test. The
closure free-variable analyzer doesn't recursively visit the
`ffi_intrinsic_call` AST node introduced in Phase 1.1, so any
identifier used inside `#objc_call` / `#jni_call` /
`#jni_static_call` from a closure body trips:
error: unresolved: '<name>'
The same identifier captured from the same scope into a plain
expression resolves fine — so the bug is localized to whatever
recursive arm-walk powers the capture analysis.
Likely fix: add an `ffi_intrinsic_call => { ... }` arm wherever
the `.call =>` arm visits `callee` + `args`. Candidate files:
- src/sema.zig (capture / scope tracking)
- src/ir/lower.zig (closure body lowering / `lowerLambda`)
Both should be checked.
Workaround in the meantime: reach the captured value via a
module-level global from inside the closure body. See the
`g_hasher_recv` pattern in
examples/ffi-objc-call-09-in-construct.sx for an applied
instance.
42 lines
1.5 KiB
Plaintext
42 lines
1.5 KiB
Plaintext
// Closure capture analysis doesn't trace into the `FfiIntrinsicCall`
|
|
// AST node — identifiers used inside `#objc_call` / `#jni_call` /
|
|
// `#jni_static_call` from a closure body aren't recognized as
|
|
// captured variables. Surfaced when writing `ffi-objc-call-09-in-
|
|
// construct.sx`.
|
|
//
|
|
// Reduced repro: capture in a closure body works fine for
|
|
// "normal" expressions (see `passthrough_works`), but the same
|
|
// capture inside `#objc_call`'s arg list trips
|
|
// "unresolved: 'recv'" (see `passthrough_via_objc_call` — would
|
|
// fail at parse time, so it's commented out).
|
|
//
|
|
// Likely fix: in the closure free-variable analyzer (sema.zig /
|
|
// lower.zig), add a recursive arm for `ffi_intrinsic_call` that
|
|
// visits `return_type` + every `args[i]` the same way the `.call`
|
|
// arm walks `callee` + `args`.
|
|
|
|
#import "modules/std.sx";
|
|
#import "modules/compiler.sx";
|
|
|
|
passthrough_works :: (recv: *void) -> Closure(s32) -> *void {
|
|
closure((d: s32) -> *void => recv); // captures `recv` — fine
|
|
}
|
|
|
|
// passthrough_via_objc_call :: (recv: *void) -> Closure(s32) -> s64 {
|
|
// // Same `recv` capture, but inside `#objc_call(...)`:
|
|
// // error: unresolved: 'recv'
|
|
// closure((d: s32) -> s64 => #objc_call(s64)(recv, "hash"));
|
|
// }
|
|
|
|
main :: () -> s32 {
|
|
inline if OS == .macos {
|
|
f := passthrough_works(null);
|
|
p := f(0);
|
|
print("ok (passthrough works) = {}\n", p == null);
|
|
}
|
|
inline if OS != .macos {
|
|
print("skipped (not macos)\n");
|
|
}
|
|
0;
|
|
}
|