embed: tests, editor support and docs; js vm file access; test fixes

embed declarations get accept and reject fixtures, a run-only image check
through zc0, an lsp keyword hover and the grammar keyword. errors point at
the declaration and still declare the name, so its uses are not reported
again. zc0 tracks byte-string globals and lays embedded arrays out one word
per element, the way its own indexing reads them.

the command-line javascript vm reaches node's file system for the file
syscalls, so include and embed work there; other embedders opt in.

the web build is self-contained in build/web with its examples. README,
BORROW and BOOTSTRAP-OWNERSHIP describe the new build and ownership rules,
and ONBOARD.md is a reading and verification order for the code.

the suite passes the map header to mklayout, the wasm fuzz program uses
the include format, and lsp drivers scale their deadlines on hosts
without the arm64 jit (ZC_TEST_TIMEOUT_SCALE). LLVM_CONFIG selects an
llvm with headers for the llvm section.

1 files changed, 16 insertions, 11 deletions

diff --git a/src/tools/zc0.c b/src/tools/zc0.c
index 5f5b97d..b0ab29f 100644
--- a/src/tools/zc0.c
+++ b/src/tools/zc0.c
@@ -202,7 +202,7 @@ static void expect(const char *s) { if (!accept(s)) { char b[64]; sprintf(b, "ex
 static char *ident(void) { static char b[8][256]; static int k; char *r = b[k++ & 7]; if (tk != T_ID) die("expected a name"); strcpy(r, ts); next(); return r; }
 
 /* ---- symbols ---- */
-typedef struct { char name[64]; int kind, sig, defd; int64_t val; } Sym;
+typedef struct { char name[64]; int kind, sig, defd, str; int64_t val; } Sym;	/* str: a global holding a byte string */
 static Sym sy[16384]; static int nsy;
 typedef struct { int at, sym; } Fix;
 static Fix fx[65536]; static int nfx;
@@ -225,7 +225,7 @@ static Sym *add(const char *n, int kind, int64_t val)
 	char full[256]; qualify(full, sizeof full, n); n = full;
 	Sym *s = find_raw(n);
 	if (s && s->kind == K_FN && kind == K_FN && !s->defd) { s->val = val; return s; }	/* prototype or forward use */
-	s = &sy[nsy++]; strncpy(s->name, n, 63); s->kind = kind; s->val = val; s->defd = 1; s->sig = 0; return s;
+	s = &sy[nsy++]; strncpy(s->name, n, 63); s->kind = kind; s->val = val; s->defd = 1; s->sig = 0; s->str = 0; return s;
 }
 static void fnref(const char *n)	/* push a function address, fixing it up later if needed */
 {
@@ -488,7 +488,7 @@ static void primary(void)
 	if ((l = findl(n))) { laddr(l->slot); lv = !l->arr; if (!l->arr) { xf = l->flt; xs = l->str; lvi = (int)(l - lo); } return; }
 	s = find(n);
 	if (s && s->kind == K_CONST) { lit(s->val); return; }
-	if (s && s->kind == K_GVAR) { lit(s->val); lv = 1; return; }
+	if (s && s->kind == K_GVAR) { lit(s->val); lv = 1; xs = s->str; return; }
 	if (s && s->kind == K_GARR) { lit(s->val); return; }
 	if (is("(")) {
 		int na = args();
@@ -799,8 +799,11 @@ static void enhdef(void)
 /* embed "path" as [type] name; at the top level: a global holding the file's bytes,
    read when the image is built. the path is relative to the working directory.
    bstr (the default) packs the bytes; char* keeps one character to a word; a
-   number takes them little-endian; T[] and T[n] lay them out as the array's
-   elements. the same rules as the resident compiler's kw_embed. */
+   number takes them little-endian; T[] and T[n] have one element per file chunk
+   of T's size. the checks are the resident compiler's (kw_embed). zc0 drops
+   types, so its arrays are one word per element whatever T is: an element is
+   stored as the value it reads as (sign extended for i8, i16 and i32, widened
+   for f32), and zc0 code indexing the array sees the same values. */
 static int embed_width(const char *t, int star)
 {
 	if (star) return 0;
@@ -832,7 +835,7 @@ static void embeddecl(void)
 	if (!buf || fread(buf, 1, (size_t)len, f) != (size_t)len) die("can't read the file named by embed");
 	fclose(f);
 	if (arr == -2 && (!ty[0] || (!strcmp(ty, "bstr") && !star))) {
-		a = here; add(n, K_GVAR, a); emit(0); img[a - GORG] = here; emit(len);
+		a = here; add(n, K_GVAR, a)->str = 1; emit(0); img[a - GORG] = here; emit(len);
 		for (i = 0; i < len; i += 8) { uint64_t u = 0; for (w = 0; w < 8 && i + w < len; w++) u |= (uint64_t)buf[i + w] << (w * 8); emit((int64_t)u); }
 		if (!(len & 7)) emit(0);
 	} else if (arr == -2 && !strcmp(ty, "char") && star == 1) {
@@ -857,12 +860,14 @@ static void embeddecl(void)
 		if (arr >= 0 && count > arr) die("the embedded file is larger than the array it is declared as");
 		if (arr >= 0) count = arr;
 		a = here;
-		if (!strcmp(ty, "char")) { for (i = 0; i < count; i++) emit(i < len ? buf[i] : 0); }
-		else {
-			long bytes = (long)count * w;
-			for (i = 0; i < bytes; i += 8) { uint64_t u = 0; for (int k = 0; k < 8 && i + k < bytes; k++) u |= (uint64_t)(i + k < len ? buf[i + k] : 0) << (k * 8); emit((int64_t)u); }
-			if (!bytes) emit(0);
+		for (i = 0; i < count; i++) {
+			uint64_t u = 0; int k;
+			for (k = 0; k < w && (long)i * w + k < len; k++) u |= (uint64_t)buf[(long)i * w + k] << (k * 8);
+			if (ty[0] == 'i' && w < 8 && (u >> (w * 8 - 1) & 1)) u |= ~(uint64_t)0 << (w * 8);
+			if (!strcmp(ty, "f32")) u = fop_s2f(u);
+			emit((int64_t)u);
 		}
+		if (!count) emit(0);
 		add(n, K_GARR, a);
 	}
 	free(buf);