john/cell
1
0
Fork 0
Code Issues 31 Pull Requests Actions Packages Projects Releases 2 Wiki Activity

Merge branch 'mach' into bytecode_cleanup

Browse Source
This commit is contained in:
John Alanbrook
2026-02-12 07:50:09 -06:00
parent 7036cdf2d1 e7a2f16004
commit 68fb440502
23 changed files with 1867 additions and 4331 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
dump_mcode.cm Normal file
View File

@@ -0,0 +1,20 @@
var fd = use("fd")
var json = use("json")
var tokenize = use("tokenize")
var parse = use("parse")
var fold = use("fold")
var mcode = use("mcode")
var streamline = use("streamline")
var name = args[0]
var src = text(fd.slurp(name))
var tok = tokenize(src, name)
var ast = parse(tok.tokens, src, name, tokenize)
var folded = fold(ast)
var compiled = mcode(folded)
var optimized = streamline(compiled)
var out = json.encode(optimized)
var f = fd.open("/tmp/mcode_dump.json", "w")
fd.write(f, out)
fd.close(f)
print("wrote /tmp/mcode_dump.json")

BIN
fold.mach
View File

Binary file not shown.

BIN
fold_new.mach Normal file
View File

Binary file not shown.

61
internal/bootstrap.cm
View File

@@ -1,5 +1,5 @@
// Hidden vars come from env:
// CLI mode (cell_init): os, args, core_path, shop_path, use_mcode
// CLI mode (cell_init): os, args, core_path, shop_path, emit_qbe
// Actor spawn (script_startup): os, json, nota, wota, actorsym, init, core_path, shop_path
// args[0] = script name, args[1..] = user args
var load_internal = os.load_internal
@@ -24,7 +24,7 @@ function use_basic(path) {
return result
}
// Load a module from .mach bytecode, falling back to .ast.json
// Load a module from .mach bytecode (bootstrap modules have no source fallback)
function boot_load(name, env) {
var mach_path = core_path + '/' + name + ".mach"
var data = null
@@ -32,9 +32,8 @@ function boot_load(name, env) {
data = fd.slurp(mach_path)
return mach_load(data, env)
}
var ast_path = core_path + '/' + name + ".ast.json"
data = text(fd.slurp(ast_path))
return mach_eval_ast(name, data, env)
print("error: missing bootstrap bytecode: " + mach_path + "\n")
disrupt
}
var boot_env = {use: use_basic}
@@ -45,14 +44,11 @@ use_cache['tokenize'] = tokenize_mod
use_cache['parse'] = parse_mod
use_cache['fold'] = fold_mod
// Optionally load mcode compiler module
var mcode_mod = null
// Always load mcode compiler module
var mcode_mod = boot_load("mcode", boot_env)
use_cache['mcode'] = mcode_mod
var streamline_mod = null
var qbe_emit_mod = null
if (use_mcode) {
mcode_mod = boot_load("mcode", boot_env)
use_cache['mcode'] = mcode_mod
}
// Warn if any .cm source is newer than its .mach bytecode
function check_mach_stale() {
@@ -134,6 +130,8 @@ function load_module(name, env) {
var src_path = null
var src = null
var ast = null
var compiled = null
var optimized = null
if (fd.is_file(mach_path)) {
data = fd.slurp(mach_path)
return mach_load(data, env)
@@ -141,38 +139,33 @@ function load_module(name, env) {
src_path = core_path + '/' + name + ".cm"
src = text(fd.slurp(src_path))
ast = analyze(src, src_path)
return mach_eval_ast(name, json.encode(ast), env)
compiled = mcode_mod(ast)
optimized = streamline_mod(compiled)
return mach_eval_mcode(name, json.encode(optimized), env)
}
// Load optimization pipeline modules (needs analyze to be defined)
var qbe_macros = null
if (use_mcode) {
streamline_mod = load_module("streamline", boot_env)
use_cache['streamline'] = streamline_mod
if (emit_qbe) {
qbe_macros = load_module("qbe", boot_env)
qbe_emit_mod = load_module("qbe_emit", boot_env)
use_cache['qbe'] = qbe_macros
use_cache['qbe_emit'] = qbe_emit_mod
}
streamline_mod = load_module("streamline", boot_env)
use_cache['streamline'] = streamline_mod
if (emit_qbe) {
qbe_macros = load_module("qbe", boot_env)
qbe_emit_mod = load_module("qbe_emit", boot_env)
use_cache['qbe'] = qbe_macros
use_cache['qbe_emit'] = qbe_emit_mod
}
// Run AST through either mcode or mach pipeline
// Run AST through mcode pipeline → register VM
function run_ast(name, ast, env) {
var compiled = null
var optimized = null
var compiled = mcode_mod(ast)
var optimized = streamline_mod(compiled)
var qbe_il = null
if (use_mcode) {
compiled = mcode_mod(ast)
optimized = streamline_mod(compiled)
if (emit_qbe) {
qbe_il = qbe_emit_mod(optimized, qbe_macros)
print(qbe_il)
return null
}
return mcode_run(name, json.encode(optimized), env)
if (emit_qbe) {
qbe_il = qbe_emit_mod(optimized, qbe_macros)
print(qbe_il)
return null
}
return mach_eval_ast(name, json.encode(ast), env)
return mach_eval_mcode(name, json.encode(optimized), env)
}
// use() with ƿit pipeline for .cm modules

BIN
internal/bootstrap.mach
View File

Binary file not shown.

BIN
internal/engine.mach
View File

Binary file not shown.

BIN
mcode.mach
View File

Binary file not shown.

1
meson.build
View File

@@ -46,7 +46,6 @@ src += [ # core
'miniz.c',
'runtime.c',
'mach.c',
'mcode.c',
'libregexp.c', 'libunicode.c', 'cutils.c', 'dtoa.c'
]

14
parse.cm
View File

@@ -65,7 +65,7 @@ var parse = function(tokens, src, filename, tokenizer) {
var errors = []
var error_count = 0
var function_nr = 1
var fn_counter = 1
var ast_node = function(kind, token) {
return {
@@ -422,8 +422,8 @@ var parse = function(tokens, src, filename, tokenizer) {
_control_depth = meth_old_cd
_control_type = meth_old_ct
_expecting_body = meth_old_eb
fn.function_nr = function_nr
function_nr = function_nr + 1
fn.function_nr = fn_counter
fn_counter = fn_counter + 1
ast_node_end(fn)
pair.right = fn
} else if (is_ident && (tok.kind == "," || tok.kind == "}")) {
@@ -918,8 +918,8 @@ var parse = function(tokens, src, filename, tokenizer) {
_control_depth = old_cd
_control_type = old_ct
_expecting_body = old_eb
node.function_nr = function_nr
function_nr = function_nr + 1
node.function_nr = fn_counter
fn_counter = fn_counter + 1
ast_node_end(node)
return node
}
@@ -1008,8 +1008,8 @@ var parse = function(tokens, src, filename, tokenizer) {
_control_depth = old_cd
_control_type = old_ct
_expecting_body = old_eb
node.function_nr = function_nr
function_nr = function_nr + 1
node.function_nr = fn_counter
fn_counter = fn_counter + 1
ast_node_end(node)
return node
}

BIN
parse.mach
View File

Binary file not shown.

BIN
qbe.mach
View File

Binary file not shown.

2
qbe_emit.cm
View File

@@ -405,7 +405,7 @@ var qbe_emit = function(ir, qbe) {
else if (op == "le_float") fop_id = 3
else if (op == "gt_float") fop_id = 4
else if (op == "ge_float") fop_id = 5
emit(qbe.cmp_float != null ? cmp_float(p, "%ctx", s(a2), s(a3), fop_id) : ` %${p} =l call $qbe_float_cmp(l %ctx, w ${text(fop_id)}, l ${s(a2)}, l ${s(a3)})`)
emit(qbe.cmp_float != null ? qbe.cmp_float(p, "%ctx", s(a2), s(a3), fop_id) : ` %${p} =l call $qbe_float_cmp(l %ctx, w ${text(fop_id)}, l ${s(a2)}, l ${s(a3)})`)
emit(` ${s(a1)} =l copy %${p}`)
wb(a1)
continue

BIN
qbe_emit.mach
View File

Binary file not shown.

36
regen.cm
View File

@@ -6,6 +6,8 @@ var json = use("json")
var tokenize = use("tokenize")
var parse = use("parse")
var fold = use("fold")
var mcode = use("mcode")
var streamline = use("streamline")
var files = [
{src: "tokenize.cm", name: "tokenize", out: "tokenize.mach"},
@@ -25,21 +27,49 @@ var src = null
var tok_result = null
var ast = null
var folded = null
var ast_json = null
var compiled = null
var optimized = null
var mcode_json = null
var bytecode = null
var f = null
var errs = null
var ei = 0
var e = null
var had_errors = false
while (i < length(files)) {
entry = files[i]
src = text(fd.slurp(entry.src))
tok_result = tokenize(src, entry.src)
ast = parse(tok_result.tokens, src, entry.src, tokenize)
// Check for parse/semantic errors
errs = ast.errors
if (errs != null && length(errs) > 0) {
ei = 0
while (ei < length(errs)) {
e = errs[ei]
if (e.line != null) {
print(`${entry.src}:${text(e.line)}:${text(e.column)}: error: ${e.message}`)
} else {
print(`${entry.src}: error: ${e.message}`)
}
ei = ei + 1
}
had_errors = true
i = i + 1
continue
}
folded = fold(ast)
ast_json = json.encode(folded)
bytecode = mach_compile_ast(entry.name, ast_json)
compiled = mcode(folded)
optimized = streamline(compiled)
mcode_json = json.encode(optimized)
bytecode = mach_compile_mcode_bin(entry.name, mcode_json)
f = fd.open(entry.out, "w")
fd.write(f, bytecode)
fd.close(f)
print(`wrote ${entry.out}`)
i = i + 1
}
if (had_errors) {
print("regen aborted: fix errors above")
}

49
source/cell.c
View File

@@ -12,7 +12,6 @@
#include "cJSON.h"
#define BOOTSTRAP_MACH "internal/bootstrap.mach"
#define BOOTSTRAP_AST "internal/bootstrap.ast.json"
#define BOOTSTRAP_SRC "internal/bootstrap.cm"
#define CELL_SHOP_DIR ".cell"
#define CELL_CORE_DIR "packages/core"
@@ -179,14 +178,9 @@ void script_startup(cell_rt *prt)
cell_rt *crt = JS_GetContextOpaque(js);
JS_FreeValue(js, js_blob_use(js));
// Load pre-compiled bootstrap bytecode (.mach), fall back to AST JSON
// Load pre-compiled bootstrap bytecode (.mach)
size_t boot_size;
int boot_is_bin = 1;
char *boot_data = load_core_file(BOOTSTRAP_MACH, &boot_size);
if (!boot_data) {
boot_is_bin = 0;
boot_data = load_core_file(BOOTSTRAP_AST, &boot_size);
}
if (!boot_data) {
printf("ERROR: Could not load bootstrap from %s!\n", core_path);
return;
@@ -211,9 +205,8 @@ void script_startup(cell_rt *prt)
JS_SetPropertyStr(js, hidden_env, "init", JS_NULL);
}
// Set args and use_mcode to null/false for actor spawn (not CLI mode)
// Set args to null for actor spawn (not CLI mode)
JS_SetPropertyStr(js, hidden_env, "args", JS_NULL);
JS_SetPropertyStr(js, hidden_env, "use_mcode", JS_NewBool(js, 0));
if (core_path)
JS_SetPropertyStr(js, hidden_env, "core_path", JS_NewString(js, core_path));
@@ -225,17 +218,8 @@ void script_startup(cell_rt *prt)
// Run through MACH VM
crt->state = ACTOR_RUNNING;
JSValue v;
if (boot_is_bin) {
v = JS_RunMachBin(js, (const uint8_t *)boot_data, boot_size, hidden_env);
free(boot_data);
} else {
cJSON *ast = cJSON_Parse(boot_data);
free(boot_data);
if (!ast) { printf("ERROR: Failed to parse bootstrap AST\n"); return; }
v = JS_RunMachTree(js, ast, hidden_env);
cJSON_Delete(ast);
}
JSValue v = JS_RunMachBin(js, (const uint8_t *)boot_data, boot_size, hidden_env);
free(boot_data);
uncaught_exception(js, v);
crt->state = ACTOR_IDLE;
set_actor_state(crt);
@@ -290,7 +274,7 @@ static void print_usage(const char *prog)
printf("Options:\n");
printf(" --core <path> Set core path directly (overrides CELL_CORE)\n");
printf(" --shop <path> Set shop path (overrides CELL_SHOP)\n");
printf(" --mcode <script> [args] Run through mcode compilation pipeline\n");
printf(" --emit-qbe Emit QBE IL (for native compilation)\n");
printf(" --test [heap_size] Run C test suite\n");
printf(" -h, --help Show this help message\n");
printf("\nEnvironment:\n");
@@ -322,7 +306,6 @@ int cell_init(int argc, char **argv)
}
/* Default: run script through bootstrap pipeline */
int use_mcode = 0;
int emit_qbe = 0;
int arg_start = 1;
const char *shop_override = NULL;
@@ -331,10 +314,9 @@ int cell_init(int argc, char **argv)
// Parse flags (order-independent)
while (arg_start < argc && argv[arg_start][0] == '-') {
if (strcmp(argv[arg_start], "--mcode") == 0) {
use_mcode = 1;
/* --mcode is now always on; accept and ignore for compat */
arg_start++;
} else if (strcmp(argv[arg_start], "--emit-qbe") == 0) {
use_mcode = 1; // QBE requires mcode pipeline
emit_qbe = 1;
arg_start++;
} else if (strcmp(argv[arg_start], "--shop") == 0) {
@@ -361,12 +343,7 @@ int cell_init(int argc, char **argv)
actor_initialize();
size_t boot_size;
int boot_is_bin = 1;
char *boot_data = load_core_file(BOOTSTRAP_MACH, &boot_size);
if (!boot_data) {
boot_is_bin = 0;
boot_data = load_core_file(BOOTSTRAP_AST, &boot_size);
}
if (!boot_data) {
printf("ERROR: Could not load bootstrap from %s\n", core_path);
return 1;
@@ -420,7 +397,6 @@ int cell_init(int argc, char **argv)
JS_SetPropertyStr(ctx, hidden_env, "core_path", JS_NewString(ctx, core_path));
JS_SetPropertyStr(ctx, hidden_env, "shop_path",
shop_path ? JS_NewString(ctx, shop_path) : JS_NULL);
JS_SetPropertyStr(ctx, hidden_env, "use_mcode", JS_NewBool(ctx, use_mcode));
JS_SetPropertyStr(ctx, hidden_env, "emit_qbe", JS_NewBool(ctx, emit_qbe));
JS_SetPropertyStr(ctx, hidden_env, "actorsym", JS_DupValue(ctx, cli_rt->actor_sym_ref.val));
JS_SetPropertyStr(ctx, hidden_env, "json", js_json_use(ctx));
@@ -435,17 +411,8 @@ int cell_init(int argc, char **argv)
JS_SetPropertyStr(ctx, hidden_env, "args", args_arr);
hidden_env = JS_Stone(ctx, hidden_env);
JSValue result;
if (boot_is_bin) {
result = JS_RunMachBin(ctx, (const uint8_t *)boot_data, boot_size, hidden_env);
free(boot_data);
} else {
cJSON *ast = cJSON_Parse(boot_data);
free(boot_data);
if (!ast) { printf("Failed to parse bootstrap AST\n"); JS_FreeContext(ctx); JS_FreeRuntime(g_runtime); return 1; }
result = JS_RunMachTree(ctx, ast, hidden_env);
cJSON_Delete(ast);
}
JSValue result = JS_RunMachBin(ctx, (const uint8_t *)boot_data, boot_size, hidden_env);
free(boot_data);
int exit_code = 0;
if (JS_IsException(result)) {

3523
source/mach.c
View File

File diff suppressed because it is too large Load Diff

2029
source/mcode.c
View File

File diff suppressed because it is too large Load Diff

220
source/quickjs-internal.h
View File

@@ -406,6 +406,8 @@ typedef struct { uint16_t line; uint16_t col; } MachLineEntry;
#define MACH_GET_sJ(i) ((int32_t)((i) & 0xFFFFFF00) >> 8)
typedef enum MachOpcode {
/* === Legacy opcodes (used by existing .mach files) === */
/* Constants & Loading */
MACH_LOADK, /* R(A) = K(Bx) — load from constant pool (ABx) */
MACH_LOADI, /* R(A) = (int16_t)sBx — load small integer (AsBx) */
@@ -416,7 +418,7 @@ typedef enum MachOpcode {
/* Movement */
MACH_MOVE, /* R(A) = R(B) */
/* Arithmetic (ABC) */
/* Generic arithmetic (ABC) — used by legacy .mach */
MACH_ADD, /* R(A) = R(B) + R(C) */
MACH_SUB, /* R(A) = R(B) - R(C) */
MACH_MUL, /* R(A) = R(B) * R(C) */
@@ -427,7 +429,7 @@ typedef enum MachOpcode {
MACH_INC, /* R(A) = R(B) + 1 */
MACH_DEC, /* R(A) = R(B) - 1 */
/* Comparison (ABC) */
/* Generic comparison (ABC) — used by legacy .mach */
MACH_EQ, /* R(A) = (R(B) == R(C)) */
MACH_NEQ, /* R(A) = (R(B) != R(C)) */
MACH_LT, /* R(A) = (R(B) < R(C)) */
@@ -435,7 +437,7 @@ typedef enum MachOpcode {
MACH_GT, /* R(A) = (R(B) > R(C)) */
MACH_GE, /* R(A) = (R(B) >= R(C)) */
/* Logical/Bitwise */
/* Logical/Bitwise — used by legacy .mach */
MACH_LNOT, /* R(A) = !R(B) */
MACH_BNOT, /* R(A) = ~R(B) */
MACH_BAND, /* R(A) = R(B) & R(C) */
@@ -445,7 +447,7 @@ typedef enum MachOpcode {
MACH_SHR, /* R(A) = R(B) >> R(C) */
MACH_USHR, /* R(A) = R(B) >>> R(C) */
/* Property access */
/* Property access — used by legacy .mach */
MACH_GETFIELD, /* R(A) = R(B)[K(C)] — named property */
MACH_SETFIELD, /* R(A)[K(B)] = R(C) — named property */
MACH_GETINDEX, /* R(A) = R(B)[R(C)] — computed property */
@@ -466,12 +468,12 @@ typedef enum MachOpcode {
MACH_JMPFALSE, /* if !R(A): pc += sBx — (iAsBx format) */
MACH_JMPNULL, /* if R(A)==null: pc += sBx */
/* Function calls — Lua-style consecutive registers */
/* Function calls — Lua-style consecutive registers (legacy .mach) */
MACH_CALL, /* Call R(A) with B args R(A+1)..R(A+B), C=0 discard, C=1 keep result in R(A) */
MACH_RETURN, /* Return R(A) */
MACH_RETNIL, /* Return null */
/* Object/array creation */
/* Object/array creation — legacy .mach */
MACH_NEWOBJECT, /* R(A) = {} */
MACH_NEWARRAY, /* R(A) = new array, B = element count in R(A+1)..R(A+B) */
MACH_CLOSURE, /* R(A) = closure(functions[Bx]) (ABx) */
@@ -493,10 +495,112 @@ typedef enum MachOpcode {
MACH_NOP,
/* === New mcode-derived opcodes (1:1 mapping to mcode IR) === */
/* Typed integer arithmetic (ABC) */
MACH_ADD_INT, /* R(A) = R(B) + R(C) — int, overflow → float */
MACH_SUB_INT, /* R(A) = R(B) - R(C) — int */
MACH_MUL_INT, /* R(A) = R(B) * R(C) — int */
MACH_DIV_INT, /* R(A) = R(B) / R(C) — int */
MACH_MOD_INT, /* R(A) = R(B) % R(C) — int */
MACH_NEG_INT, /* R(A) = -R(B) — int (AB) */
/* Typed float arithmetic (ABC) */
MACH_ADD_FLOAT, /* R(A) = R(B) + R(C) — float */
MACH_SUB_FLOAT, /* R(A) = R(B) - R(C) — float */
MACH_MUL_FLOAT, /* R(A) = R(B) * R(C) — float */
MACH_DIV_FLOAT, /* R(A) = R(B) / R(C) — float */
MACH_MOD_FLOAT, /* R(A) = R(B) % R(C) — float */
MACH_NEG_FLOAT, /* R(A) = -R(B) — float (AB) */
/* Text */
MACH_CONCAT, /* R(A) = R(B) ++ R(C) — string concatenation */
/* Typed integer comparisons (ABC) */
MACH_EQ_INT, /* R(A) = (R(B) == R(C)) — int */
MACH_NE_INT, /* R(A) = (R(B) != R(C)) — int */
MACH_LT_INT, /* R(A) = (R(B) < R(C)) — int */
MACH_LE_INT, /* R(A) = (R(B) <= R(C)) — int */
MACH_GT_INT, /* R(A) = (R(B) > R(C)) — int */
MACH_GE_INT, /* R(A) = (R(B) >= R(C)) — int */
/* Typed float comparisons (ABC) */
MACH_EQ_FLOAT, /* R(A) = (R(B) == R(C)) — float */
MACH_NE_FLOAT, /* R(A) = (R(B) != R(C)) — float */
MACH_LT_FLOAT, /* R(A) = (R(B) < R(C)) — float */
MACH_LE_FLOAT, /* R(A) = (R(B) <= R(C)) — float */
MACH_GT_FLOAT, /* R(A) = (R(B) > R(C)) — float */
MACH_GE_FLOAT, /* R(A) = (R(B) >= R(C)) — float */
/* Typed text comparisons (ABC) */
MACH_EQ_TEXT, /* R(A) = (R(B) == R(C)) — text */
MACH_NE_TEXT, /* R(A) = (R(B) != R(C)) — text */
MACH_LT_TEXT, /* R(A) = (R(B) < R(C)) — text */
MACH_LE_TEXT, /* R(A) = (R(B) <= R(C)) — text */
MACH_GT_TEXT, /* R(A) = (R(B) > R(C)) — text */
MACH_GE_TEXT, /* R(A) = (R(B) >= R(C)) — text */
/* Typed bool comparisons (ABC) */
MACH_EQ_BOOL, /* R(A) = (R(B) == R(C)) — bool */
MACH_NE_BOOL, /* R(A) = (R(B) != R(C)) — bool */
/* Special comparisons */
MACH_IS_IDENTICAL, /* R(A) = (R(B) === R(C)) — identity check (ABC) */
/* Type checks (AB) */
MACH_IS_INT, /* R(A) = is_int(R(B)) */
MACH_IS_NUM, /* R(A) = is_num(R(B)) */
MACH_IS_TEXT, /* R(A) = is_text(R(B)) */
MACH_IS_BOOL, /* R(A) = is_bool(R(B)) */
MACH_IS_NULL, /* R(A) = is_null(R(B)) */
MACH_TYPEOF, /* R(A) = typeof(R(B)) */
/* Logical (mcode-style) */
MACH_NOT, /* R(A) = !R(B) — boolean not (AB) */
MACH_AND, /* R(A) = R(B) && R(C) (ABC) */
MACH_OR, /* R(A) = R(B) || R(C) (ABC) */
/* Bitwise (mcode names) */
MACH_BITNOT, /* R(A) = ~R(B) (AB) */
MACH_BITAND, /* R(A) = R(B) & R(C) (ABC) */
MACH_BITOR, /* R(A) = R(B) | R(C) (ABC) */
MACH_BITXOR, /* R(A) = R(B) ^ R(C) (ABC) */
/* Property access (mcode names) */
MACH_LOAD_FIELD, /* R(A) = R(B).K(C) — named property (ABC) */
MACH_STORE_FIELD, /* R(A).K(B) = R(C) — named property (ABC) */
MACH_LOAD_INDEX, /* R(A) = R(B)[R(C)] — integer index (ABC) */
MACH_STORE_INDEX, /* R(A)[R(B)] = R(C) — integer index (ABC) */
MACH_LOAD_DYNAMIC, /* R(A) = R(B)[R(C)] — dynamic key (ABC) */
MACH_STORE_DYNAMIC, /* R(A)[R(B)] = R(C) — dynamic key (ABC) */
/* Object/Array creation (mcode names) */
MACH_NEWRECORD, /* R(A) = {} — new empty record (A only) */
/* Decomposed function calls (mcode-style) */
MACH_FRAME, /* R(A) = frame(R(B), C) — alloc call frame (ABC) */
MACH_SETARG, /* frame R(A)[B] = R(C) — set arg in frame (ABC) */
MACH_INVOKE, /* R(B) = invoke(R(A)) — call frame, result in R(B) (AB) */
MACH_GOFRAME, /* R(A) = goframe(R(B), C) — async frame (ABC) */
MACH_GOINVOKE, /* goinvoke(R(A)) — async invoke, no result (A only) */
/* Control flow */
MACH_JMPNOTNULL, /* if R(A)!=null: pc += sBx (iAsBx) */
/* Error handling */
MACH_DISRUPT, /* trigger disruption (A only) */
/* Variable storage */
MACH_SET_VAR, /* env/global[K(Bx)] = R(A) — store to var (ABx) */
/* Misc */
MACH_IN, /* R(A) = (R(B) in R(C)) — has property (ABC) */
MACH_OP_COUNT
} MachOpcode;
static const char *mach_opcode_names[MACH_OP_COUNT] = {
/* Legacy */
[MACH_LOADK] = "loadk",
[MACH_LOADI] = "loadi",
[MACH_LOADNULL] = "loadnull",
@@ -556,6 +660,70 @@ static const char *mach_opcode_names[MACH_OP_COUNT] = {
[MACH_EQ_TOL] = "eq_tol",
[MACH_NEQ_TOL] = "neq_tol",
[MACH_NOP] = "nop",
/* Mcode-derived */
[MACH_ADD_INT] = "add_int",
[MACH_SUB_INT] = "sub_int",
[MACH_MUL_INT] = "mul_int",
[MACH_DIV_INT] = "div_int",
[MACH_MOD_INT] = "mod_int",
[MACH_NEG_INT] = "neg_int",
[MACH_ADD_FLOAT] = "add_float",
[MACH_SUB_FLOAT] = "sub_float",
[MACH_MUL_FLOAT] = "mul_float",
[MACH_DIV_FLOAT] = "div_float",
[MACH_MOD_FLOAT] = "mod_float",
[MACH_NEG_FLOAT] = "neg_float",
[MACH_CONCAT] = "concat",
[MACH_EQ_INT] = "eq_int",
[MACH_NE_INT] = "ne_int",
[MACH_LT_INT] = "lt_int",
[MACH_LE_INT] = "le_int",
[MACH_GT_INT] = "gt_int",
[MACH_GE_INT] = "ge_int",
[MACH_EQ_FLOAT] = "eq_float",
[MACH_NE_FLOAT] = "ne_float",
[MACH_LT_FLOAT] = "lt_float",
[MACH_LE_FLOAT] = "le_float",
[MACH_GT_FLOAT] = "gt_float",
[MACH_GE_FLOAT] = "ge_float",
[MACH_EQ_TEXT] = "eq_text",
[MACH_NE_TEXT] = "ne_text",
[MACH_LT_TEXT] = "lt_text",
[MACH_LE_TEXT] = "le_text",
[MACH_GT_TEXT] = "gt_text",
[MACH_GE_TEXT] = "ge_text",
[MACH_EQ_BOOL] = "eq_bool",
[MACH_NE_BOOL] = "ne_bool",
[MACH_IS_IDENTICAL] = "is_identical",
[MACH_IS_INT] = "is_int",
[MACH_IS_NUM] = "is_num",
[MACH_IS_TEXT] = "is_text",
[MACH_IS_BOOL] = "is_bool",
[MACH_IS_NULL] = "is_null",
[MACH_TYPEOF] = "typeof",
[MACH_NOT] = "not",
[MACH_AND] = "and",
[MACH_OR] = "or",
[MACH_BITNOT] = "bitnot",
[MACH_BITAND] = "bitand",
[MACH_BITOR] = "bitor",
[MACH_BITXOR] = "bitxor",
[MACH_LOAD_FIELD] = "load_field",
[MACH_STORE_FIELD] = "store_field",
[MACH_LOAD_INDEX] = "load_index",
[MACH_STORE_INDEX] = "store_index",
[MACH_LOAD_DYNAMIC] = "load_dynamic",
[MACH_STORE_DYNAMIC] = "store_dynamic",
[MACH_NEWRECORD] = "newrecord",
[MACH_FRAME] = "frame",
[MACH_SETARG] = "setarg",
[MACH_INVOKE] = "invoke",
[MACH_GOFRAME] = "goframe",
[MACH_GOINVOKE] = "goinvoke",
[MACH_JMPNOTNULL] = "jmpnotnull",
[MACH_DISRUPT] = "disrupt",
[MACH_SET_VAR] = "set_var",
[MACH_IN] = "in",
};
/* Compiled register-based code (off-heap, never GC'd).
@@ -586,28 +754,6 @@ typedef struct JSCodeRegister {
uint16_t disruption_pc; /* start of disruption handler (0 = none) */
} JSCodeRegister;
/* Pre-parsed MCODE for a single function (off-heap, never GC'd).
Created by jsmcode_parse from cJSON MCODE output.
Instructions remain as cJSON pointers for string-based dispatch. */
typedef struct JSMCode {
uint16_t nr_args;
uint16_t nr_slots;
/* Pre-flattened instruction array (cJSON array items → C array for O(1) access) */
cJSON **instrs;
uint32_t instr_count;
/* Label map: label string → instruction index */
struct { const char *name; uint32_t index; } *labels;
uint32_t label_count;
/* Nested function definitions (indexes into top-level functions array) */
struct JSMCode **functions;
uint32_t func_count;
/* Keep root cJSON alive (owns all the cJSON nodes instrs[] point into) */
cJSON *json_root;
MachLineEntry *line_table; /* [instr_count], parallel to instrs[] */
const char *name; /* function name (points into cJSON tree) */
const char *filename; /* source filename (points into cJSON tree) */
uint16_t disruption_pc; /* start of disruption handler (0 = none) */
} JSMCode;
/* Frame for closures - used by link-time relocation model where closures
reference outer frames via (depth, slot) addressing.
@@ -1068,7 +1214,6 @@ typedef enum {
JS_FUNC_KIND_BYTECODE,
JS_FUNC_KIND_C_DATA,
JS_FUNC_KIND_REGISTER, /* register-based VM function */
JS_FUNC_KIND_MCODE, /* MCODE JSON interpreter */
} JSFunctionKind;
typedef struct JSFunction {
@@ -1092,11 +1237,6 @@ typedef struct JSFunction {
JSValue env_record; /* stone record, module environment */
JSValue outer_frame; /* JSFrame JSValue, for closures */
} reg;
struct {
JSMCode *code; /* pre-parsed MCODE (off-heap) */
JSValue outer_frame; /* lexical parent frame for closures */
JSValue env_record; /* module env or JS_NULL */
} mcode;
} u;
} JSFunction;
@@ -1259,8 +1399,6 @@ JSValue js_call_c_function (JSContext *ctx, JSValue func_obj, JSValue this_obj,
JSValue js_call_bound_function (JSContext *ctx, JSValue func_obj, JSValue this_obj, int argc, JSValue *argv);
JSValue JS_CallInternal (JSContext *ctx, JSValue func_obj, JSValue this_obj, int argc, JSValue *argv, int flags);
JSValue JS_CallRegisterVM(JSContext *ctx, JSCodeRegister *code, JSValue this_obj, int argc, JSValue *argv, JSValue env, JSValue outer_frame);
JSValue mcode_exec(JSContext *ctx, JSMCode *code, JSValue this_obj,
int argc, JSValue *argv, JSValue outer_frame);
int JS_DeleteProperty (JSContext *ctx, JSValue obj, JSValue prop);
JSValue __attribute__ ((format (printf, 2, 3)))
JS_ThrowInternalError (JSContext *ctx, const char *fmt, ...);
@@ -1574,13 +1712,6 @@ typedef struct {
} GetLineColCache;
/* === MachVarInfo (shared by mach.c and mcode.c) === */
typedef struct MachVarInfo {
char *name;
int slot;
int is_const; /* 1 for def, function args; 0 for var */
int is_closure; /* 1 if captured by a nested function */
} MachVarInfo;
/* === PPretext (parser pretext, system-malloc, used by cell_js.c parser) === */
typedef struct PPretext {
@@ -1770,8 +1901,5 @@ JSFrameRegister *alloc_frame_register(JSContext *ctx, int slot_count);
cJSON *mach_find_scope_record(cJSON *scopes, int function_nr);
int reg_vm_check_interrupt(JSContext *ctx);
/* mcode.c exports */
JSValue mcode_exec(JSContext *ctx, JSMCode *code, JSValue this_obj, int argc, JSValue *argv, JSValue outer_frame);
#endif /* QUICKJS_INTERNAL_H */

23
source/quickjs.h
View File

@@ -992,30 +992,11 @@ MachCode *JS_DeserializeMachCode(const uint8_t *data, size_t size);
/* Load compiled MachCode into a JSContext, materializing JSValues. */
struct JSCodeRegister *JS_LoadMachCode(JSContext *ctx, MachCode *mc, JSValue env);
/* Dump MACH bytecode to stdout. Takes AST cJSON tree. */
void JS_DumpMachTree (JSContext *ctx, struct cJSON *ast, JSValue env);
/* Dump MACH bytecode to stdout. Takes AST JSON string. */
void JS_DumpMach (JSContext *ctx, const char *ast_json, JSValue env);
/* Compile and execute MACH bytecode from AST cJSON tree. */
JSValue JS_RunMachTree (JSContext *ctx, struct cJSON *ast, JSValue env);
/* Compile and execute MACH bytecode from AST JSON string. */
JSValue JS_RunMach (JSContext *ctx, const char *ast_json, JSValue env);
/* Deserialize and execute pre-compiled MACH binary bytecode. */
JSValue JS_RunMachBin(JSContext *ctx, const uint8_t *data, size_t size, JSValue env);
/* Execute MCODE from cJSON tree. Takes ownership of root. */
JSValue JS_CallMcodeTree (JSContext *ctx, struct cJSON *root);
/* Execute MCODE from cJSON tree with hidden env. Takes ownership of root. */
JSValue JS_CallMcodeTreeEnv (JSContext *ctx, struct cJSON *root, JSValue env);
/* Parse and execute MCODE JSON string.
Returns result of execution, or JS_EXCEPTION on error. */
JSValue JS_CallMcode (JSContext *ctx, const char *mcode_json);
/* Compile mcode JSON IR to MachCode binary. */
MachCode *mach_compile_mcode(struct cJSON *mcode_json);
/* Get stack trace as cJSON array of frame objects.
Returns NULL if no register VM frame is active.

178
source/runtime.c
View File

@@ -1143,10 +1143,6 @@ void gc_scan_object (JSContext *ctx, void *ptr, uint8_t *from_base, uint8_t *fro
nested->name = gc_copy_value (ctx, nested->name, from_base, from_end, to_base, to_free, to_end);
}
}
} else if (fn->kind == JS_FUNC_KIND_MCODE) {
/* MCODE function - scan outer_frame and env_record */
fn->u.mcode.outer_frame = gc_copy_value (ctx, fn->u.mcode.outer_frame, from_base, from_end, to_base, to_free, to_end);
fn->u.mcode.env_record = gc_copy_value (ctx, fn->u.mcode.env_record, from_base, from_end, to_base, to_free, to_end);
}
break;
}
@@ -4336,8 +4332,6 @@ JSValue JS_CallInternal (JSContext *ctx, JSValue func_obj, JSValue this_obj,
case JS_FUNC_KIND_REGISTER:
return JS_CallRegisterVM (ctx, f->u.reg.code, this_obj, argc, argv,
f->u.reg.env_record, f->u.reg.outer_frame);
case JS_FUNC_KIND_MCODE:
return mcode_exec (ctx, f->u.mcode.code, this_obj, argc, argv, f->u.mcode.outer_frame);
default:
return JS_ThrowTypeError (ctx, "not a function");
}
@@ -4359,8 +4353,6 @@ JSValue JS_Call (JSContext *ctx, JSValue func_obj, JSValue this_obj, int argc, J
case JS_FUNC_KIND_REGISTER:
return JS_CallRegisterVM (ctx, f->u.reg.code, this_obj, argc, argv,
f->u.reg.env_record, f->u.reg.outer_frame);
case JS_FUNC_KIND_MCODE:
return mcode_exec (ctx, f->u.mcode.code, this_obj, argc, argv, f->u.mcode.outer_frame);
default:
return JS_ThrowTypeError (ctx, "not a function");
}
@@ -9721,83 +9713,6 @@ int js_is_blob (JSContext *js, JSValue v) {
* ============================================================================
*/
/* mach_eval_ast(name, ast_json, env?) - compile pre-parsed AST and run */
static JSValue js_mach_eval_ast (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
if (argc < 2 || !JS_IsText (argv[0]) || !JS_IsText (argv[1]))
return JS_ThrowTypeError (ctx, "mach_eval_ast requires (name, ast_json) text arguments");
const char *name = JS_ToCString (ctx, argv[0]);
if (!name) return JS_EXCEPTION;
const char *json_str = JS_ToCString (ctx, argv[1]);
if (!json_str) {
JS_FreeCString (ctx, name);
return JS_EXCEPTION;
}
cJSON *ast = cJSON_Parse (json_str);
JS_FreeCString (ctx, json_str);
if (!ast) {
JS_FreeCString (ctx, name);
return JS_ThrowSyntaxError (ctx, "mach_eval_ast: failed to parse AST JSON");
}
/* Set the filename on the AST root */
cJSON_DeleteItemFromObjectCaseSensitive (ast, "filename");
cJSON_AddStringToObject (ast, "filename", name);
JSValue env = (argc >= 3 && JS_IsGCObject (argv[2])) ? argv[2] : JS_NULL;
JSValue result = JS_RunMachTree (ctx, ast, env);
cJSON_Delete (ast);
JS_FreeCString (ctx, name);
return result;
}
/* mach_compile_ast(name, ast_json) - compile pre-parsed AST to binary blob */
static JSValue js_mach_compile_ast (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
if (argc < 2 || !JS_IsText (argv[0]) || !JS_IsText (argv[1]))
return JS_ThrowTypeError (ctx, "mach_compile_ast requires (name, ast_json) text arguments");
const char *name = JS_ToCString (ctx, argv[0]);
if (!name) return JS_EXCEPTION;
const char *json_str = JS_ToCString (ctx, argv[1]);
if (!json_str) {
JS_FreeCString (ctx, name);
return JS_EXCEPTION;
}
cJSON *ast = cJSON_Parse (json_str);
JS_FreeCString (ctx, json_str);
if (!ast) {
JS_FreeCString (ctx, name);
return JS_ThrowSyntaxError (ctx, "mach_compile_ast: failed to parse AST JSON");
}
cJSON_DeleteItemFromObjectCaseSensitive (ast, "filename");
cJSON_AddStringToObject (ast, "filename", name);
MachCode *mc = JS_CompileMachTree (ast);
cJSON_Delete (ast);
JS_FreeCString (ctx, name);
if (!mc)
return JS_ThrowSyntaxError (ctx, "mach_compile_ast: failed to compile AST");
size_t blob_size;
uint8_t *buf = JS_SerializeMachCode (mc, &blob_size);
JS_FreeMachCode (mc);
if (!buf)
return JS_ThrowInternalError (ctx, "mach_compile_ast: serialization failed");
JSValue result = js_new_blob_stoned_copy (ctx, buf, blob_size);
sys_free (buf);
return result;
}
/* mach_load(blob, env?) - deserialize and execute binary blob */
static JSValue js_mach_load (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
if (argc < 1)
@@ -9824,10 +9739,10 @@ static JSValue js_mach_load (JSContext *ctx, JSValue this_val, int argc, JSValue
return result;
}
/* mcode_run(name, mcode_json, env?) - run pre-compiled mcode JSON */
static JSValue js_mcode_run (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
/* mach_eval_mcode(name, mcode_json, env?) - compile mcode IR and run via register VM */
static JSValue js_mach_eval_mcode (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
if (argc < 2 || !JS_IsText (argv[0]) || !JS_IsText (argv[1]))
return JS_ThrowTypeError (ctx, "mcode_run requires (name, mcode_json) text arguments");
return JS_ThrowTypeError (ctx, "mach_eval_mcode requires (name, mcode_json) text arguments");
const char *name = JS_ToCString (ctx, argv[0]);
if (!name) return JS_EXCEPTION;
@@ -9843,16 +9758,81 @@ static JSValue js_mcode_run (JSContext *ctx, JSValue this_val, int argc, JSValue
if (!mcode) {
JS_FreeCString (ctx, name);
return JS_ThrowSyntaxError (ctx, "mcode_run: failed to parse mcode JSON");
return JS_ThrowSyntaxError (ctx, "mach_eval_mcode: failed to parse mcode JSON");
}
/* Set filename on the mcode root if not present */
if (!cJSON_GetObjectItemCaseSensitive (mcode, "filename"))
cJSON_AddStringToObject (mcode, "filename", name);
/* Compile mcode IR → MachCode binary */
MachCode *mc = mach_compile_mcode (mcode);
cJSON_Delete (mcode);
if (!mc) {
JS_FreeCString (ctx, name);
return JS_ThrowInternalError (ctx, "mach_eval_mcode: compilation failed");
}
JSValue env = (argc >= 3 && JS_IsGCObject (argv[2])) ? argv[2] : JS_NULL;
JSValue result = JS_CallMcodeTreeEnv (ctx, mcode, env);
/* mcode tree ownership transferred to JS_CallMcodeTreeEnv — do not free */
JSGCRef env_ref;
JS_PushGCRef (ctx, &env_ref);
env_ref.val = env;
JSCodeRegister *code = JS_LoadMachCode (ctx, mc, env_ref.val);
JS_FreeMachCode (mc);
JSValue result = JS_CallRegisterVM (ctx, code, ctx->global_obj, 0, NULL, env_ref.val, JS_NULL);
JS_PopGCRef (ctx, &env_ref);
JS_FreeCString (ctx, name);
return result;
}
/* mach_compile_mcode_bin(name, mcode_json) - compile mcode IR to serialized binary blob */
static JSValue js_mach_compile_mcode_bin (JSContext *ctx, JSValue this_val, int argc, JSValue *argv) {
if (argc < 2 || !JS_IsText (argv[0]) || !JS_IsText (argv[1]))
return JS_ThrowTypeError (ctx, "mach_compile_mcode_bin requires (name, mcode_json) text arguments");
const char *name = JS_ToCString (ctx, argv[0]);
if (!name) return JS_EXCEPTION;
const char *json_str = JS_ToCString (ctx, argv[1]);
if (!json_str) {
JS_FreeCString (ctx, name);
return JS_EXCEPTION;
}
cJSON *mcode = cJSON_Parse (json_str);
JS_FreeCString (ctx, json_str);
if (!mcode) {
JS_FreeCString (ctx, name);
return JS_ThrowSyntaxError (ctx, "mach_compile_mcode_bin: failed to parse mcode JSON");
}
if (!cJSON_GetObjectItemCaseSensitive (mcode, "filename"))
cJSON_AddStringToObject (mcode, "filename", name);
MachCode *mc = mach_compile_mcode (mcode);
cJSON_Delete (mcode);
JS_FreeCString (ctx, name);
if (!mc)
return JS_ThrowInternalError (ctx, "mach_compile_mcode_bin: compilation failed");
size_t size = 0;
uint8_t *data = JS_SerializeMachCode (mc, &size);
JS_FreeMachCode (mc);
if (!data)
return JS_ThrowInternalError (ctx, "mach_compile_mcode_bin: serialization failed");
JSValue blob = js_new_blob_stoned_copy (ctx, data, size);
sys_free (data);
return blob;
}
/* ============================================================================
* stone() function - deep freeze with blob support
* ============================================================================
@@ -10938,10 +10918,9 @@ static void JS_AddIntrinsicBaseObjects (JSContext *ctx) {
}
/* Core functions - using GC-safe helper */
js_set_global_cfunc(ctx, "mach_eval_ast", js_mach_eval_ast, 3);
js_set_global_cfunc(ctx, "mcode_run", js_mcode_run, 3);
js_set_global_cfunc(ctx, "mach_compile_ast", js_mach_compile_ast, 2);
js_set_global_cfunc(ctx, "mach_load", js_mach_load, 2);
js_set_global_cfunc(ctx, "mach_eval_mcode", js_mach_eval_mcode, 3);
js_set_global_cfunc(ctx, "mach_compile_mcode_bin", js_mach_compile_mcode_bin, 2);
js_set_global_cfunc(ctx, "stone", js_cell_stone, 1);
js_set_global_cfunc(ctx, "length", js_cell_length, 1);
js_set_global_cfunc(ctx, "call", js_cell_call, 3);
@@ -12326,17 +12305,6 @@ cJSON *JS_GetStack(JSContext *ctx) {
line = code->line_table[pc].line;
col = code->line_table[pc].col;
}
} else if (fn->kind == JS_FUNC_KIND_MCODE && fn->u.mcode.code) {
JSMCode *code = fn->u.mcode.code;
file = code->filename;
func_name = code->name;
if (!is_first) {
pc = (uint32_t)(JS_VALUE_GET_INT(frame->address) >> 16);
}
if (code->line_table && pc < code->instr_count) {
line = code->line_table[pc].line;
col = code->line_table[pc].col;
}
}
cJSON *entry = cJSON_CreateObject();

BIN
streamline.mach
View File

Binary file not shown.

42
test_parse_boot.cm Normal file
View File

@@ -0,0 +1,42 @@
var fd = use("fd")
var json = use("json")
var tokenize = use("tokenize")
var parse = use("parse")
var fold = use("fold")
var src = text(fd.slurp("internal/bootstrap.cm"))
var tok_result = tokenize(src, "bootstrap.cm")
var ast = parse(tok_result.tokens, src, "bootstrap.cm", tokenize)
var i = 0
var folded = null
var ast_json = null
var f = null
var bytecode = null
var has_errors = ast.errors != null && length(ast.errors) > 0
if (has_errors) {
print("PARSE ERRORS:")
while (i < length(ast.errors)) {
print(text(ast.errors[i].line) + ":" + text(ast.errors[i].column) + " " + ast.errors[i].message)
i = i + 1
}
} else {
print("Parse OK")
print(" statements: " + text(length(ast.statements)))
if (ast.functions != null) {
print(" functions: " + text(length(ast.functions)))
} else {
print(" functions: null")
}
folded = fold(ast)
ast_json = json.encode(folded)
f = fd.open("/tmp/bootstrap_ast.json", "w")
fd.write(f, ast_json)
fd.close(f)
print("Wrote AST to /tmp/bootstrap_ast.json")
bytecode = mach_compile_ast("bootstrap", ast_json)
print("Bytecode size: " + text(length(bytecode)))
f = fd.open("/tmp/bootstrap_test.mach", "w")
fd.write(f, bytecode)
fd.close(f)
print("Wrote bytecode to /tmp/bootstrap_test.mach")
}

BIN
tokenize.mach
View File

Binary file not shown.