Hojun-Cho 368b85ea33 cmd: α/γ-4 rt.malloc → alloc([], N)! (w6a/w6c/w6l file slurps + sized bufs) ...

Phase 0 #8 fourth α-batch. 10 α/γ conversions across 6 files:
 - w6a/main.ww slurp (1)
 - w6c/main.ww slurp (1)
 - w6l/dyn.ww slurpso (1)
 - w6l/obj.ww slurp (1)
 - w6l/dynout.ww filebuf at :661 (1; threads .ptr through ~50
   dwr*/dbcopy callees — that's the 139-line w6l/dynout diff)
 - w6l/out.ww hdr at :92 (1)
 - w6l/main.ww 4 sites: bufp + γ {inputs, libdirs, lflags} (4)

Deferred:
 - w6l/main.ww 3× 16/8/20-byte δ fixed-bufs at :67/:106/:176 →
   stack-promote in a separate sub-task.
 - w6l/obj.ww :113/:135 β text-buffer growth → memio.dynamic refactor.

Pattern unchanged from 47918d3/00d88ff/e9eb67d: `alloc([], N: u64)!` +
`buf.len = N: i32;` + `buf.ptr` for callees taking `*u8`. γ uses
element-count (was bytes): `maxinputs: u64` not `maxinputs*8u64`.

Verified 132/132 + 995_self_rebuild byte-identity. Advances #45.

2026-05-21 01:59:58 +09:00

23 KiB

Raw Blame History

View Raw