lib/rt: rename rt_alloc → rt_malloc; rt.alloc → rt.malloc

Hare's canonical runtime allocator is rt::malloc with linker symbol
rt.malloc (ref/hare/rt/malloc.ha:27,78). ww kept the dot→underscore
Plan 9 convention (CLAUDE.md rule 4) so the linker symbol becomes
rt_malloc; the lib/rt exported function name becomes malloc; ww
callers say rt.malloc(...).

The language builtin keyword stays `alloc(T)!` — unchanged from Hare
(ref/hare/hare/lex/token.ha:21 ltok::ALLOC, parse/expr.ha:398
builtin()). The rename only touches the lowered linker symbol and the
exported function name behind it; the user-facing syntax for
heap-allocation is identical to Hare.

Surface:
- rt/alloc.s: TEXT rt_alloc → TEXT rt_malloc, labels updated
- lib/rt/malloc.ww: @symbol("rt_malloc") fn malloc(...) (was rt_alloc/alloc)
- rt/ensure.ww: local FFI decl + call site updated to malloc; `!` dropped
  on the direct FFI call (rt_malloc returns *void, not a tagged union)
- 18 .ww callers: rt.alloc(...) → rt.malloc(...)
- cstage cmd/wcc/check.c + wwstage selfhost/cmd/wcc/check.ww
  alloc-builtin suppression gate routes through ffi_resolve("malloc")
  for the lowering; the user-shadow check still keys on the BUILTIN
  KEYWORD "alloc" since that is what `alloc(...)` parses as. Adding
  "malloc" to the user-shadow check was unnecessary and was reverted
  during pre-commit review.
- cstage cmd/w6c/cgen.c: 2× ffi_resolve("alloc") → ffi_resolve("malloc")
- wwstage cgenexpr/cgenstmt: 2× ffiresolve(c, "alloc") → ffiresolve(c, "malloc")
- Test fixtures (700_e2e, 758_cgalloc_str_field, 990_selfhost, 992_w6l_ww,
  selfhost/test/tagged_ptr_ret.ww): updated inline ww sources to the new
  decl + call form

This is commit 2 of 3 in the lib/rt extraction (#38). Commit 3 closes
the OOM contract — return type becomes nullable *void and the builtin
lowering null-checks + propagates nomem.

Verified 132/132 + 995_self_rebuild byte-identity (5 wwstage tools
round-trip identical) + make clean cold rebuild.

lib/strings/strings.ww

@@ -99,7 +99,7 @@ export fn dup(s: str) str = {
 	r.ptr = nil;
 	r.len = 0;
 	if (s.len == 0) { return r; };
-	let buf: *u8 = rt.alloc(s.len: u64): *u8;
+	let buf: *u8 = rt.malloc(s.len: u64): *u8;
 	let i: i32 = 0;
 	for (i < s.len) { buf[i] = s[i]; i += 1; };
 	r.ptr = buf;
@@ -121,7 +121,7 @@ export fn dup(s: str) str = {
 // (#46). The pre-allocated slice has `cap == s.len`, so appendstr's
 // rt_ensure call never reaches the grow branch.
 //
-// Empty input bypasses the alloc: rt_alloc(0) is an mmap of 0 bytes
+// Empty input bypasses the alloc: rt_malloc(0) is an mmap of 0 bytes
 // which returns -EINVAL, and the alloc-slice `?` shortcut routes
 // that through nomem — Hare's heap allocator hands back a sentinel
 // instead (#47). Return `{nil, 0, 0}` directly so callers get the
@@ -177,7 +177,7 @@ export fn concat(strs: str...) str = {
 	r.ptr = nil;
 	r.len = 0;
 	if (total == 0) { return r; };
-	let buf: *u8 = rt.alloc(total: u64): *u8;
+	let buf: *u8 = rt.malloc(total: u64): *u8;
 	let off: i32 = 0;
 	i = 0;
 	for (i < strs.len) {
@@ -210,7 +210,7 @@ export fn join(delim: str, strs: str...) str = {
 	r.ptr = nil;
 	r.len = 0;
 	if (total == 0) { return r; };
-	let buf: *u8 = rt.alloc(total: u64): *u8;
+	let buf: *u8 = rt.malloc(total: u64): *u8;
 	let off: i32 = 0;
 	i = 0;
 	for (i < strs.len) {
@@ -887,7 +887,7 @@ export fn lpad(s: str, p: rune, maxlen: i32) str = {
 	if (s.len >= maxlen) { return dup(s); };
 	let scratch: [4]u8;
 	let pad: []u8 = runebytes(scratch[0:4], p);
-	let buf: *u8 = rt.alloc(maxlen: u64): *u8;
+	let buf: *u8 = rt.malloc(maxlen: u64): *u8;
 	let padwrite: i32 = (maxlen - s.len) * pad.len;
 	if (padwrite > maxlen) { padwrite = maxlen; };
 	let off: i32 = 0;
@@ -970,7 +970,7 @@ export fn rpad(s: str, p: rune, maxlen: i32) str = {
 	if (s.len >= maxlen) { return dup(s); };
 	let scratch: [4]u8;
 	let pad: []u8 = runebytes(scratch[0:4], p);
-	let buf: *u8 = rt.alloc(maxlen: u64): *u8;
+	let buf: *u8 = rt.malloc(maxlen: u64): *u8;
 	let k: i32 = 0;
 	for (k < s.len) {
 		buf[k] = s[k];