Add support for embedding LuaJIT bytecode for builtins.

12 files changed, 199 insertions, 31 deletions

diff --git a/src/Makefile b/src/Makefile
index 4ea8c85e..f3631a0d 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -558,6 +558,10 @@ amalg:
 clean:
         $(HOST_RM) $(ALL_RM)
 
+libbc:
+        ./$(LUAJIT_T) host/genlibbc.lua $(LJLIB_C) >host/buildvm_libbc.h
+        $(MAKE) all
+
 depend:
         @for file in $(ALL_HDRGEN); do \
           test -f $$file || touch $$file; \
@@ -572,7 +576,7 @@ depend:
           test -s $$file || $(HOST_RM) $$file; \
           done
 
-.PHONY: default all amalg clean depend
+.PHONY: default all amalg clean libbc depend
 
 ##############################################################################
 # Rules for generated files.
diff --git a/src/Makefile.dep b/src/Makefile.dep
index 5d91723a..54004f7c 100644
--- a/src/Makefile.dep
+++ b/src/Makefile.dep
@@ -124,7 +124,8 @@ lj_lex.o: lj_lex.c lj_obj.h lua.h luaconf.h lj_def.h lj_arch.h lj_gc.h \
  lj_state.h lj_lex.h lj_parse.h lj_char.h lj_strscan.h
 lj_lib.o: lj_lib.c lauxlib.h lua.h luaconf.h lj_obj.h lj_def.h lj_arch.h \
  lj_gc.h lj_err.h lj_errmsg.h lj_str.h lj_tab.h lj_func.h lj_bc.h \
- lj_dispatch.h lj_jit.h lj_ir.h lj_vm.h lj_strscan.h lj_lib.h
+ lj_dispatch.h lj_jit.h lj_ir.h lj_vm.h lj_strscan.h lj_lex.h lj_bcdump.h \
+ lj_lib.h
 lj_load.o: lj_load.c lua.h luaconf.h lauxlib.h lj_obj.h lj_def.h \
  lj_arch.h lj_gc.h lj_err.h lj_errmsg.h lj_str.h lj_func.h lj_frame.h \
  lj_bc.h lj_vm.h lj_lex.h lj_bcdump.h lj_parse.h
@@ -220,7 +221,8 @@ host/buildvm_asm.o: host/buildvm_asm.c host/buildvm.h lj_def.h lua.h luaconf.h \
 host/buildvm_fold.o: host/buildvm_fold.c host/buildvm.h lj_def.h lua.h \
  luaconf.h lj_arch.h lj_obj.h lj_def.h lj_arch.h lj_ir.h lj_obj.h
 host/buildvm_lib.o: host/buildvm_lib.c host/buildvm.h lj_def.h lua.h luaconf.h \
- lj_arch.h lj_obj.h lj_def.h lj_arch.h lj_lib.h lj_obj.h
+ lj_arch.h lj_obj.h lj_def.h lj_arch.h lj_lib.h lj_obj.h \
+ host/buildvm_libbc.h
 host/buildvm_peobj.o: host/buildvm_peobj.c host/buildvm.h lj_def.h lua.h \
  luaconf.h lj_arch.h lj_bc.h lj_def.h lj_arch.h
 host/minilua.o: host/minilua.c
diff --git a/src/host/buildvm_lib.c b/src/host/buildvm_lib.c
index 40141dfb..182ab90f 100644
--- a/src/host/buildvm_lib.c
+++ b/src/host/buildvm_lib.c
@@ -6,6 +6,7 @@
 #include "buildvm.h"
 #include "lj_obj.h"
 #include "lj_lib.h"
+#include "buildvm_libbc.h"
 
 /* Context for library definitions. */
 static uint8_t obuf[8192];
@@ -151,6 +152,55 @@ static void libdef_func(BuildCtx *ctx, char *p, int arg)
   regfunc = REGFUNC_OK;
 }
 
+static uint32_t libdef_uleb128(uint8_t **pp)
+{
+  uint8_t *p = *pp;
+  uint32_t v = *p++;
+  if (v >= 0x80) {
+    int sh = 0; v &= 0x7f;
+    do { v |= ((*p & 0x7f) << (sh += 7)); } while (*p++ >= 0x80);
+  }
+  *pp = p;
+  return v;
+}
+
+static void libdef_swapbc(uint8_t *p)
+{
+  uint32_t i, sizebc;
+  p += 4;
+  libdef_uleb128(&p);
+  libdef_uleb128(&p);
+  sizebc = libdef_uleb128(&p);
+  for (i = 0; i < sizebc; i++, p += 4) {
+    uint8_t t = p[0]; p[0] = p[3]; p[3] = t;
+    t = p[1]; p[1] = p[2]; p[2] = t;
+  }
+}
+
+static void libdef_lua(BuildCtx *ctx, char *p, int arg)
+{
+  UNUSED(arg);
+  if (ctx->mode == BUILD_libdef) {
+    int i;
+    for (i = 0; libbc_map[i].name != NULL; i++) {
+      if (!strcmp(libbc_map[i].name, p)) {
+        int ofs = libbc_map[i].ofs;
+        int len = libbc_map[i+1].ofs - ofs;
+        obuf[2]++;  /* Bump hash table size. */
+        *optr++ = LIBINIT_LUA;
+        libdef_name(p, 0);
+        memcpy(optr, libbc_code + ofs, len);
+        if (libbc_endian != LJ_BE)
+          libdef_swapbc(optr);
+        optr += len;
+        return;
+      }
+    }
+    fprintf(stderr, "Error: missing libbc definition for %s\n", p);
+    exit(1);
+  }
+}
+
 static uint32_t find_rec(char *name)
 {
   char *p = (char *)obuf;
@@ -277,6 +327,7 @@ static const LibDefHandler libdef_handlers[] = {
   { "CF(",      ")",            libdef_func,            LIBINIT_CF },
   { "ASM(",     ")",            libdef_func,            LIBINIT_ASM },
   { "ASM_(",    ")",            libdef_func,            LIBINIT_ASM_ },
+  { "LUA(",     ")",            libdef_lua,             0 },
   { "REC(",     ")",            libdef_rec,             0 },
   { "PUSH(",    ")",            libdef_push,            0 },
   { "SET(",     ")",            libdef_set,             0 },
diff --git a/src/host/buildvm_libbc.h b/src/host/buildvm_libbc.h
new file mode 100644
index 00000000..d2d83ea6
--- /dev/null
+++ b/src/host/buildvm_libbc.h
@@ -0,0 +1,12 @@
+/* This is a generated file. DO NOT EDIT! */
+
+static const int libbc_endian = 0;
+
+static const uint8_t libbc_code[] = {
+0
+};
+
+static const struct { const char *name; int ofs; } libbc_map[] = {
+{NULL,0}
+};
+
diff --git a/src/host/genlibbc.lua b/src/host/genlibbc.lua
new file mode 100644
index 00000000..b0dbf17a
--- /dev/null
+++ b/src/host/genlibbc.lua
@@ -0,0 +1,68 @@
+----------------------------------------------------------------------------
+-- Lua script to dump the bytecode of the library functions written in Lua.
+-- The resulting 'buildvm_libbc.h' is used for the build process of LuaJIT.
+----------------------------------------------------------------------------
+-- Copyright (C) 2005-2013 Mike Pall. All rights reserved.
+-- Released under the MIT license. See Copyright Notice in luajit.h
+----------------------------------------------------------------------------
+
+local function usage()
+  io.stderr:write("Usage: ", arg and arg[0] or "genlibbc", " lib_*.c\n")
+  os.exit(1)
+end
+
+local function read_source()
+  if not (arg and arg[1]) then usage() end
+  local src = ""
+  for _,name in ipairs(arg) do
+    local fp = assert(io.open(name))
+    src = src .. fp:read("*a")
+    fp:close()
+  end
+  return src
+end
+
+local function find_defs(src)
+  local defs = {}
+  for name, code in string.gmatch(src, "LJLIB_LUA%(([^)]*)%)%s*/%*(.-)%*/") do
+    local env = {}
+    local func = assert(load("return "..code, "", nil, env))()
+    local d = string.dump(func, true)
+    local ofs = 6
+    while string.byte(d, ofs) > 127 do ofs = ofs + 1 end
+    defs[name] = string.sub(d, ofs+1, -2)
+    defs[#defs+1] = name
+  end
+  return defs
+end
+
+local function write_defs(fp, defs)
+  fp:write("/* This is a generated file. DO NOT EDIT! */\n\n")
+  fp:write("static const int libbc_endian = ",
+           string.byte(string.dump(function() end), 5) % 2, ";\n\n")
+  local s = ""
+  for _,name in ipairs(defs) do
+    s = s .. defs[name]
+  end
+  fp:write("static const uint8_t libbc_code[] = {\n")
+  local n = 0
+  for i=1,#s do
+    local x = string.byte(s, i)
+    fp:write(x, ",")
+    n = n + (x < 10 and 2 or (x < 100 and 3 or 4))
+    if n >= 75 then n = 0; fp:write("\n") end
+  end
+  fp:write("0\n};\n\n")
+  fp:write("static const struct { const char *name; int ofs; } libbc_map[] = {\n")
+  local m = 0
+  for _,name in ipairs(defs) do
+    fp:write('{"', name, '",', m, '},\n')
+    m = m + #defs[name]
+  end
+  fp:write("{NULL,", m, "}\n};\n\n")
+  fp:flush()
+end
+
+local src = read_source()
+local defs = find_defs(src)
+write_defs(io.stdout, defs)
diff --git a/src/lj_bcdump.h b/src/lj_bcdump.h
index e660156d..c1ed54e7 100644
--- a/src/lj_bcdump.h
+++ b/src/lj_bcdump.h
@@ -61,6 +61,7 @@ enum {
 
 LJ_FUNC int lj_bcwrite(lua_State *L, GCproto *pt, lua_Writer writer,
                        void *data, int strip);
+LJ_FUNC GCproto *lj_bcread_proto(LexState *ls);
 LJ_FUNC GCproto *lj_bcread(LexState *ls);
 
 #endif
diff --git a/src/lj_bcread.c b/src/lj_bcread.c
index 2b5ba855..7a8c08f5 100644
--- a/src/lj_bcread.c
+++ b/src/lj_bcread.c
@@ -326,25 +326,13 @@ static void bcread_uv(LexState *ls, GCproto *pt, MSize sizeuv)
 }
 
 /* Read a prototype. */
-static GCproto *bcread_proto(LexState *ls)
+GCproto *lj_bcread_proto(LexState *ls)
 {
   GCproto *pt;
   MSize framesize, numparams, flags, sizeuv, sizekgc, sizekn, sizebc, sizept;
   MSize ofsk, ofsuv, ofsdbg;
   MSize sizedbg = 0;
   BCLine firstline = 0, numline = 0;
-  MSize len, startn;
-
-  /* Read length. */
-  if (ls->n > 0 && ls->p[0] == 0) {  /* Shortcut EOF. */
-    ls->n--; ls->p++;
-    return NULL;
-  }
-  bcread_want(ls, 5);
-  len = bcread_uleb128(ls);
-  if (!len) return NULL;  /* EOF */
-  bcread_need(ls, len);
-  startn = ls->n;
 
   /* Read prototype header. */
   flags = bcread_byte(ls);
@@ -413,9 +401,6 @@ static GCproto *bcread_proto(LexState *ls)
     setmref(pt->uvinfo, NULL);
     setmref(pt->varinfo, NULL);
   }
-
-  if (len != startn - ls->n)
-    bcread_error(ls, LJ_ERR_BCBAD);
   return pt;
 }
 
@@ -462,8 +447,21 @@ GCproto *lj_bcread(LexState *ls)
   if (!bcread_header(ls))
     bcread_error(ls, LJ_ERR_BCFMT);
   for (;;) {  /* Process all prototypes in the bytecode dump. */
-    GCproto *pt = bcread_proto(ls);
-    if (!pt) break;
+    GCproto *pt;
+    MSize len, startn;
+    /* Read length. */
+    if (ls->n > 0 && ls->p[0] == 0) {  /* Shortcut EOF. */
+      ls->n--; ls->p++;
+      break;
+    }
+    bcread_want(ls, 5);
+    len = bcread_uleb128(ls);
+    if (!len) break;  /* EOF */
+    bcread_need(ls, len);
+    startn = ls->n;
+    pt = lj_bcread_proto(ls);
+    if (len != startn - ls->n)
+      bcread_error(ls, LJ_ERR_BCBAD);
     setprotoV(L, L->top, pt);
     incr_top(L);
   }
diff --git a/src/lj_debug.c b/src/lj_debug.c
index be7fb2b1..ec56b7d2 100644
--- a/src/lj_debug.c
+++ b/src/lj_debug.c
@@ -321,7 +321,7 @@ const char *lj_debug_funcname(lua_State *L, TValue *frame, const char **name)
 /* -- Source code locations ----------------------------------------------- */
 
 /* Generate shortened source name. */
-void lj_debug_shortname(char *out, GCstr *str)
+void lj_debug_shortname(char *out, GCstr *str, BCLine line)
 {
   const char *src = strdata(str);
   if (*src == '=') {
@@ -335,11 +335,11 @@ void lj_debug_shortname(char *out, GCstr *str)
       *out++ = '.'; *out++ = '.'; *out++ = '.';
     }
     strcpy(out, src);
-  } else {  /* Output [string "string"]. */
+  } else {  /* Output [string "string"] or [builtin:name]. */
     size_t len;  /* Length, up to first control char. */
     for (len = 0; len < LUA_IDSIZE-12; len++)
       if (((const unsigned char *)src)[len] < ' ') break;
-    strcpy(out, "[string \""); out += 9;
+    strcpy(out, line == ~(BCLine)0 ? "[builtin:" : "[string \""); out += 9;
     if (src[len] != '\0') {  /* Must truncate? */
       if (len > LUA_IDSIZE-15) len = LUA_IDSIZE-15;
       strncpy(out, src, len); out += len;
@@ -347,7 +347,7 @@ void lj_debug_shortname(char *out, GCstr *str)
     } else {
       strcpy(out, src); out += len;
     }
-    strcpy(out, "\"]");
+    strcpy(out, line == ~(BCLine)0 ? "]" : "\"]");
   }
 }
 
@@ -360,8 +360,9 @@ void lj_debug_addloc(lua_State *L, const char *msg,
     if (isluafunc(fn)) {
       BCLine line = debug_frameline(L, fn, nextframe);
       if (line >= 0) {
+        GCproto *pt = funcproto(fn);
         char buf[LUA_IDSIZE];
-        lj_debug_shortname(buf, proto_chunkname(funcproto(fn)));
+        lj_debug_shortname(buf, proto_chunkname(pt), pt->firstline);
         lj_str_pushf(L, "%s:%d: %s", buf, line, msg);
         return;
       }
@@ -377,7 +378,9 @@ void lj_debug_pushloc(lua_State *L, GCproto *pt, BCPos pc)
   const char *s = strdata(name);
   MSize i, len = name->len;
   BCLine line = lj_debug_line(pt, pc);
-  if (*s == '@') {
+  if (pt->firstline == ~(BCLine)0) {
+    lj_str_pushf(L, "builtin:%s", s);
+  } else if (*s == '@') {
     s++; len--;
     for (i = len; i > 0; i--)
       if (s[i] == '/' || s[i] == '\\') {
@@ -453,7 +456,7 @@ int lj_debug_getinfo(lua_State *L, const char *what, lj_Debug *ar, int ext)
         BCLine firstline = pt->firstline;
         GCstr *name = proto_chunkname(pt);
         ar->source = strdata(name);
-        lj_debug_shortname(ar->short_src, name);
+        lj_debug_shortname(ar->short_src, name, pt->firstline);
         ar->linedefined = (int)firstline;
         ar->lastlinedefined = (int)(firstline + pt->numline);
         ar->what = firstline ? "Lua" : "main";
diff --git a/src/lj_debug.h b/src/lj_debug.h
index 7cf57de7..4144b47e 100644
--- a/src/lj_debug.h
+++ b/src/lj_debug.h
@@ -34,7 +34,7 @@ LJ_FUNC const char *lj_debug_slotname(GCproto *pt, const BCIns *pc,
                                       BCReg slot, const char **name);
 LJ_FUNC const char *lj_debug_funcname(lua_State *L, TValue *frame,
                                       const char **name);
-LJ_FUNC void lj_debug_shortname(char *out, GCstr *str);
+LJ_FUNC void lj_debug_shortname(char *out, GCstr *str, BCLine line);
 LJ_FUNC void lj_debug_addloc(lua_State *L, const char *msg,
                              cTValue *frame, cTValue *nextframe);
 LJ_FUNC void lj_debug_pushloc(lua_State *L, GCproto *pt, BCPos pc);
diff --git a/src/lj_err.c b/src/lj_err.c
index 4a33a233..e0fb7167 100644
--- a/src/lj_err.c
+++ b/src/lj_err.c
@@ -587,7 +587,7 @@ LJ_NOINLINE void lj_err_lex(lua_State *L, GCstr *src, const char *tok,
 {
   char buff[LUA_IDSIZE];
   const char *msg;
-  lj_debug_shortname(buff, src);
+  lj_debug_shortname(buff, src, line);
   msg = lj_str_pushvf(L, err2msg(em), argp);
   msg = lj_str_pushf(L, "%s:%d: %s", buff, line, msg);
   if (tok)
diff --git a/src/lj_lib.c b/src/lj_lib.c
index 331eaa6a..be3ee004 100644
--- a/src/lj_lib.c
+++ b/src/lj_lib.c
@@ -18,6 +18,8 @@
 #include "lj_dispatch.h"
 #include "lj_vm.h"
 #include "lj_strscan.h"
+#include "lj_lex.h"
+#include "lj_bcdump.h"
 #include "lj_lib.h"
 
 /* -- Library initialization ---------------------------------------------- */
@@ -43,6 +45,28 @@ static GCtab *lib_create_table(lua_State *L, const char *libname, int hsize)
   return tabV(L->top-1);
 }
 
+static const uint8_t *lib_read_lfunc(lua_State *L, const uint8_t *p, GCtab *tab)
+{
+  int len = *p++;
+  GCstr *name = lj_str_new(L, (const char *)p, len);
+  LexState ls;
+  GCproto *pt;
+  GCfunc *fn;
+  memset(&ls, 0, sizeof(ls));
+  ls.L = L;
+  ls.p = (const char *)(p+len);
+  ls.n = ~(MSize)0;
+  ls.current = -1;
+  ls.level = (BCDUMP_F_STRIP|(LJ_BE*BCDUMP_F_BE));
+  ls.chunkname = name;
+  pt = lj_bcread_proto(&ls);
+  pt->firstline = ~(BCLine)0;
+  fn = lj_func_newL_empty(L, pt, tabref(L->env));
+  /* NOBARRIER: See below for common barrier. */
+  setfuncV(L, lj_tab_setstr(L, tab, name), fn);
+  return (const uint8_t *)ls.p;
+}
+
 void lj_lib_register(lua_State *L, const char *libname,
                      const uint8_t *p, const lua_CFunction *cf)
 {
@@ -87,6 +111,9 @@ void lj_lib_register(lua_State *L, const char *libname,
       ofn = fn;
     } else {
       switch (tag | len) {
+      case LIBINIT_LUA:
+        p = lib_read_lfunc(L, p, tab);
+        break;
       case LIBINIT_SET:
         L->top -= 2;
         if (tvisstr(L->top+1) && strV(L->top+1)->len == 0)
diff --git a/src/lj_lib.h b/src/lj_lib.h
index 2fe6d2a8..05f90de5 100644
--- a/src/lj_lib.h
+++ b/src/lj_lib.h
@@ -77,6 +77,7 @@ static LJ_AINLINE void lj_lib_pushcc(lua_State *L, lua_CFunction f,
 #define LJLIB_CF(name)          static int lj_cf_##name(lua_State *L)
 #define LJLIB_ASM(name)         static int lj_ffh_##name(lua_State *L)
 #define LJLIB_ASM_(name)
+#define LJLIB_LUA(name)
 #define LJLIB_SET(name)
 #define LJLIB_PUSH(arg)
 #define LJLIB_REC(handler)
@@ -96,7 +97,8 @@ LJ_FUNC void lj_lib_register(lua_State *L, const char *libname,
 #define LIBINIT_ASM     0x40
 #define LIBINIT_ASM_    0x80
 #define LIBINIT_STRING  0xc0
-#define LIBINIT_MAXSTR  0x39
+#define LIBINIT_MAXSTR  0x38
+#define LIBINIT_LUA     0xf9
 #define LIBINIT_SET     0xfa
 #define LIBINIT_NUMBER  0xfb
 #define LIBINIT_COPY    0xfc