7 files changed, 410 insertions, 26 deletions

diff --git a/src/host/buildvm.c b/src/host/buildvm.c
index b9560f31..de23fabd 100644
--- a/src/host/buildvm.c
+++ b/src/host/buildvm.c
@@ -59,10 +59,10 @@ static int collect_reloc(BuildCtx *ctx, uint8_t *addr, int idx, int type);
 #include "../dynasm/dasm_x86.h"
 #elif LJ_TARGET_ARM
 #include "../dynasm/dasm_arm.h"
+#elif LJ_TARGET_ARM64
+#include "../dynasm/dasm_arm64.h"
 #elif LJ_TARGET_PPC
 #include "../dynasm/dasm_ppc.h"
-#elif LJ_TARGET_PPCSPE
-#include "../dynasm/dasm_ppc.h"
 #elif LJ_TARGET_MIPS
 #include "../dynasm/dasm_mips.h"
 #else
@@ -110,11 +110,11 @@ static const char *sym_decorate(BuildCtx *ctx,
   if (p) {
 #if LJ_TARGET_X86ORX64
     if (!LJ_64 && (ctx->mode == BUILD_coffasm || ctx->mode == BUILD_peobj))
-      name[0] = '@';
+      name[0] = name[1] == 'R' ? '_' : '@';  /* Just for _RtlUnwind@16. */
     else
       *p = '\0';
-#elif (LJ_TARGET_PPC  || LJ_TARGET_PPCSPE) && !LJ_TARGET_CONSOLE
-    /* Keep @plt. */
+#elif LJ_TARGET_PPC && !LJ_TARGET_CONSOLE
+    /* Keep @plt etc. */
 #else
     *p = '\0';
 #endif
@@ -179,6 +179,7 @@ static int build_code(BuildCtx *ctx)
   ctx->nreloc = 0;
 
   ctx->globnames = globnames;
+  ctx->extnames = extnames;
   ctx->relocsym = (const char **)malloc(NRELOCSYM*sizeof(const char *));
   ctx->nrelocsym = 0;
   for (i = 0; i < (int)NRELOCSYM; i++) relocmap[i] = -1;
@@ -320,20 +321,20 @@ static void emit_vmdef(BuildCtx *ctx)
   char buf[80];
   int i;
   fprintf(ctx->fp, "-- This is a generated file. DO NOT EDIT!\n\n");
-  fprintf(ctx->fp, "module(...)\n\n");
+  fprintf(ctx->fp, "return {\n\n");
 
   fprintf(ctx->fp, "bcnames = \"");
   for (i = 0; bc_names[i]; i++) fprintf(ctx->fp, "%-6s", bc_names[i]);
-  fprintf(ctx->fp, "\"\n\n");
+  fprintf(ctx->fp, "\",\n\n");
 
   fprintf(ctx->fp, "irnames = \"");
   for (i = 0; ir_names[i]; i++) fprintf(ctx->fp, "%-6s", ir_names[i]);
-  fprintf(ctx->fp, "\"\n\n");
+  fprintf(ctx->fp, "\",\n\n");
 
   fprintf(ctx->fp, "irfpm = { [0]=");
   for (i = 0; irfpm_names[i]; i++)
     fprintf(ctx->fp, "\"%s\", ", lower(buf, irfpm_names[i]));
-  fprintf(ctx->fp, "}\n\n");
+  fprintf(ctx->fp, "},\n\n");
 
   fprintf(ctx->fp, "irfield = { [0]=");
   for (i = 0; irfield_names[i]; i++) {
@@ -343,17 +344,17 @@ static void emit_vmdef(BuildCtx *ctx)
     if (p) *p = '.';
     fprintf(ctx->fp, "\"%s\", ", buf);
   }
-  fprintf(ctx->fp, "}\n\n");
+  fprintf(ctx->fp, "},\n\n");
 
   fprintf(ctx->fp, "ircall = {\n[0]=");
   for (i = 0; ircall_names[i]; i++)
     fprintf(ctx->fp, "\"%s\",\n", ircall_names[i]);
-  fprintf(ctx->fp, "}\n\n");
+  fprintf(ctx->fp, "},\n\n");
 
   fprintf(ctx->fp, "traceerr = {\n[0]=");
   for (i = 0; trace_errors[i]; i++)
     fprintf(ctx->fp, "\"%s\",\n", trace_errors[i]);
-  fprintf(ctx->fp, "}\n\n");
+  fprintf(ctx->fp, "},\n\n");
 }
 
 /* -- Argument parsing ---------------------------------------------------- */
@@ -490,6 +491,7 @@ int main(int argc, char **argv)
   case BUILD_vmdef:
     emit_vmdef(ctx);
     emit_lib(ctx);
+    fprintf(ctx->fp, "}\n\n");
     break;
   case BUILD_ffdef:
   case BUILD_libdef:
diff --git a/src/host/buildvm.h b/src/host/buildvm.h
index e6dd3dcb..b90428dc 100644
--- a/src/host/buildvm.h
+++ b/src/host/buildvm.h
@@ -82,6 +82,7 @@ typedef struct BuildCtx {
   const char *beginsym;
   /* Strings generated by DynASM. */
   const char *const *globnames;
+  const char *const *extnames;
   const char *dasm_ident;
   const char *dasm_arch;
   /* Relocations. */
diff --git a/src/host/buildvm_asm.c b/src/host/buildvm_asm.c
index 81a3969a..43595b31 100644
--- a/src/host/buildvm_asm.c
+++ b/src/host/buildvm_asm.c
@@ -51,8 +51,8 @@ static const char *const jccnames[] = {
   "js", "jns", "jpe", "jpo", "jl", "jge", "jle", "jg"
 };
 
-/* Emit relocation for the incredibly stupid OSX assembler. */
-static void emit_asm_reloc_mach(BuildCtx *ctx, uint8_t *cp, int n,
+/* Emit x86/x64 text relocations. */
+static void emit_asm_reloc_text(BuildCtx *ctx, uint8_t *cp, int n,
                                 const char *sym)
 {
   const char *opname = NULL;
@@ -71,6 +71,20 @@ err:
     exit(1);
   }
   emit_asm_bytes(ctx, cp, n);
+  if (strncmp(sym+(*sym == '_'), LABEL_PREFIX, sizeof(LABEL_PREFIX)-1)) {
+    /* Various fixups for external symbols outside of our binary. */
+    if (ctx->mode == BUILD_elfasm) {
+      if (LJ_32)
+        fprintf(ctx->fp, "#if __PIC__\n\t%s lj_wrap_%s\n#else\n", opname, sym);
+      fprintf(ctx->fp, "\t%s %s@PLT\n", opname, sym);
+      if (LJ_32)
+        fprintf(ctx->fp, "#endif\n");
+      return;
+    } else if (LJ_32 && ctx->mode == BUILD_machasm) {
+      fprintf(ctx->fp, "\t%s L%s$stub\n", opname, sym);
+      return;
+    }
+  }
   fprintf(ctx->fp, "\t%s %s\n", opname, sym);
 }
 #else
@@ -79,10 +93,14 @@ static void emit_asm_words(BuildCtx *ctx, uint8_t *p, int n)
 {
   int i;
   for (i = 0; i < n; i += 4) {
+    uint32_t ins = *(uint32_t *)(p+i);
+#if LJ_TARGET_ARM64 && LJ_BE
+    ins = lj_bswap(ins);  /* ARM64 instructions are always little-endian. */
+#endif
     if ((i & 15) == 0)
-      fprintf(ctx->fp, "\t.long 0x%08x", *(uint32_t *)(p+i));
+      fprintf(ctx->fp, "\t.long 0x%08x", ins);
     else
-      fprintf(ctx->fp, ",0x%08x", *(uint32_t *)(p+i));
+      fprintf(ctx->fp, ",0x%08x", ins);
     if ((i & 15) == 12) putc('\n', ctx->fp);
   }
   if ((n & 15) != 0) putc('\n', ctx->fp);
@@ -107,7 +125,16 @@ static void emit_asm_wordreloc(BuildCtx *ctx, uint8_t *p, int n,
             ins, sym);
     exit(1);
   }
-#elif LJ_TARGET_PPC || LJ_TARGET_PPCSPE
+#elif LJ_TARGET_ARM64
+  if ((ins >> 26) == 0x25u) {
+    fprintf(ctx->fp, "\tbl %s\n", sym);
+  } else {
+    fprintf(stderr,
+            "Error: unsupported opcode %08x for %s symbol relocation.\n",
+            ins, sym);
+    exit(1);
+  }
+#elif LJ_TARGET_PPC
 #if LJ_TARGET_PS3
 #define TOCPREFIX "."
 #else
@@ -117,6 +144,14 @@ static void emit_asm_wordreloc(BuildCtx *ctx, uint8_t *p, int n,
     fprintf(ctx->fp, "\t%s %d, %d, " TOCPREFIX "%s\n",
             (ins & 1) ? "bcl" : "bc", (ins >> 21) & 31, (ins >> 16) & 31, sym);
   } else if ((ins >> 26) == 18) {
+#if LJ_ARCH_PPC64
+    const char *suffix = strchr(sym, '@');
+    if (suffix && suffix[1] == 'h') {
+      fprintf(ctx->fp, "\taddis 11, 2, %s\n", sym);
+    } else if (suffix && suffix[1] == 'l') {
+      fprintf(ctx->fp, "\tld 12, %s\n", sym);
+    } else
+#endif
     fprintf(ctx->fp, "\t%s " TOCPREFIX "%s\n", (ins & 1) ? "bl" : "b", sym);
   } else {
     fprintf(stderr,
@@ -215,6 +250,9 @@ void emit_asm(BuildCtx *ctx)
   int i, rel;
 
   fprintf(ctx->fp, "\t.file \"buildvm_%s.dasc\"\n", ctx->dasm_arch);
+#if LJ_ARCH_PPC64
+  fprintf(ctx->fp, "\t.abiversion 2\n");
+#endif
   fprintf(ctx->fp, "\t.text\n");
   emit_asm_align(ctx, 4);
 
@@ -228,11 +266,20 @@ void emit_asm(BuildCtx *ctx)
 
 #if LJ_TARGET_ARM && defined(__GNUC__) && !LJ_NO_UNWIND
   /* This should really be moved into buildvm_arm.dasc. */
+#if LJ_ARCH_HASFPU
+  fprintf(ctx->fp,
+          ".fnstart\n"
+          ".save {r5, r6, r7, r8, r9, r10, r11, lr}\n"
+          ".vsave {d8-d15}\n"
+          ".save {r4}\n"
+          ".pad #28\n");
+#else
   fprintf(ctx->fp,
           ".fnstart\n"
           ".save {r4, r5, r6, r7, r8, r9, r10, r11, lr}\n"
           ".pad #28\n");
 #endif
+#endif
 #if LJ_TARGET_MIPS
   fprintf(ctx->fp, ".set nomips16\n.abicalls\n.set noreorder\n.set nomacro\n");
 #endif
@@ -255,8 +302,9 @@ void emit_asm(BuildCtx *ctx)
       BuildReloc *r = &ctx->reloc[rel];
       int n = r->ofs - ofs;
 #if LJ_TARGET_X86ORX64
-      if (ctx->mode == BUILD_machasm && r->type != 0) {
-        emit_asm_reloc_mach(ctx, ctx->code+ofs, n, ctx->relocsym[r->sym]);
+      if (r->type != 0 &&
+          (ctx->mode == BUILD_elfasm || ctx->mode == BUILD_machasm)) {
+        emit_asm_reloc_text(ctx, ctx->code+ofs, n, ctx->relocsym[r->sym]);
       } else {
         emit_asm_bytes(ctx, ctx->code+ofs, n);
         emit_asm_reloc(ctx, r->type, ctx->relocsym[r->sym]);
@@ -290,10 +338,7 @@ void emit_asm(BuildCtx *ctx)
 #if !(LJ_TARGET_PS3 || LJ_TARGET_PSVITA)
     fprintf(ctx->fp, "\t.section .note.GNU-stack,\"\"," ELFASM_PX "progbits\n");
 #endif
-#if LJ_TARGET_PPCSPE
-    /* Soft-float ABI + SPE. */
-    fprintf(ctx->fp, "\t.gnu_attribute 4, 2\n\t.gnu_attribute 8, 3\n");
-#elif LJ_TARGET_PPC && !LJ_TARGET_PS3
+#if LJ_TARGET_PPC && !LJ_TARGET_PS3 && !LJ_ABI_SOFTFP
     /* Hard-float ABI. */
     fprintf(ctx->fp, "\t.gnu_attribute 4, 1\n");
 #endif
diff --git a/src/host/buildvm_lib.c b/src/host/buildvm_lib.c
index 569e2caf..2956fdb6 100644
--- a/src/host/buildvm_lib.c
+++ b/src/host/buildvm_lib.c
@@ -5,7 +5,9 @@
 
 #include "buildvm.h"
 #include "lj_obj.h"
+#include "lj_bc.h"
 #include "lj_lib.h"
+#include "buildvm_libbc.h"
 
 /* Context for library definitions. */
 static uint8_t obuf[8192];
@@ -151,6 +153,62 @@ static void libdef_func(BuildCtx *ctx, char *p, int arg)
   regfunc = REGFUNC_OK;
 }
 
+static uint8_t *libdef_uleb128(uint8_t *p, uint32_t *vv)
+{
+  uint32_t v = *p++;
+  if (v >= 0x80) {
+    int sh = 0; v &= 0x7f;
+    do { v |= ((*p & 0x7f) << (sh += 7)); } while (*p++ >= 0x80);
+  }
+  *vv = v;
+  return p;
+}
+
+static void libdef_fixupbc(uint8_t *p)
+{
+  uint32_t i, sizebc;
+  p += 4;
+  p = libdef_uleb128(p, &sizebc);
+  p = libdef_uleb128(p, &sizebc);
+  p = libdef_uleb128(p, &sizebc);
+  for (i = 0; i < sizebc; i++, p += 4) {
+    uint8_t op = p[libbc_endian ? 3 : 0];
+    uint8_t ra = p[libbc_endian ? 2 : 1];
+    uint8_t rc = p[libbc_endian ? 1 : 2];
+    uint8_t rb = p[libbc_endian ? 0 : 3];
+    if (!LJ_DUALNUM && op == BC_ISTYPE && rc == ~LJ_TNUMX+1) {
+      op = BC_ISNUM; rc++;
+    }
+    p[LJ_ENDIAN_SELECT(0, 3)] = op;
+    p[LJ_ENDIAN_SELECT(1, 2)] = ra;
+    p[LJ_ENDIAN_SELECT(2, 1)] = rc;
+    p[LJ_ENDIAN_SELECT(3, 0)] = rb;
+  }
+}
+
+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);
+        libdef_fixupbc(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 +335,7 @@ 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 },
@@ -373,7 +432,7 @@ void emit_lib(BuildCtx *ctx)
       "#ifndef FF_NUM_ASMFUNC\n#define FF_NUM_ASMFUNC %d\n#endif\n\n",
       ffasmfunc);
   } else if (ctx->mode == BUILD_vmdef) {
-    fprintf(ctx->fp, "}\n\n");
+    fprintf(ctx->fp, "},\n\n");
   } else if (ctx->mode == BUILD_bcdef) {
     int i;
     fprintf(ctx->fp, "\n};\n\n");
diff --git a/src/host/buildvm_libbc.h b/src/host/buildvm_libbc.h
new file mode 100644
index 00000000..b2600bd5
--- /dev/null
+++ b/src/host/buildvm_libbc.h
@@ -0,0 +1,56 @@
+/* This is a generated file. DO NOT EDIT! */
+
+static const int libbc_endian = 0;
+
+static const uint8_t libbc_code[] = {
+#if LJ_FR2
+0,1,2,0,0,1,2,24,1,0,0,76,1,2,0,241,135,158,166,3,220,203,178,130,4,0,1,2,0,
+0,1,2,24,1,0,0,76,1,2,0,243,244,148,165,20,198,190,199,252,3,0,1,2,0,0,0,3,
+16,0,5,0,21,1,0,0,76,1,2,0,0,2,10,0,0,0,15,16,0,12,0,16,1,9,0,41,2,1,0,21,3,
+0,0,41,4,1,0,77,2,8,128,18,6,1,0,18,8,5,0,59,9,5,0,66,6,3,2,10,6,0,0,88,7,1,
+128,76,6,2,0,79,2,248,127,75,0,1,0,0,2,11,0,0,0,16,16,0,12,0,16,1,9,0,43,2,
+0,0,18,3,0,0,41,4,0,0,88,5,7,128,18,7,1,0,18,9,5,0,18,10,6,0,66,7,3,2,10,7,
+0,0,88,8,1,128,76,7,2,0,70,5,3,3,82,5,247,127,75,0,1,0,0,1,2,0,0,0,3,16,0,12,
+0,21,1,0,0,76,1,2,0,0,2,10,0,0,2,30,16,0,12,0,21,2,0,0,11,1,0,0,88,3,7,128,
+8,2,0,0,88,3,23,128,59,3,2,0,43,4,0,0,64,4,2,0,76,3,2,0,88,3,18,128,16,1,14,
+0,41,3,1,0,3,3,1,0,88,3,14,128,3,1,2,0,88,3,12,128,59,3,1,0,22,4,1,1,18,5,2,
+0,41,6,1,0,77,4,4,128,23,8,1,7,59,9,7,0,64,9,8,0,79,4,252,127,43,4,0,0,64,4,
+2,0,76,3,2,0,75,0,1,0,0,2,0,5,12,0,0,0,35,16,0,12,0,16,1,14,0,16,2,14,0,16,
+3,14,0,11,4,0,0,88,5,1,128,18,4,0,0,16,4,12,0,3,1,2,0,88,5,24,128,33,5,1,3,
+0,2,3,0,88,6,4,128,2,3,1,0,88,6,2,128,4,4,0,0,88,6,9,128,18,6,1,0,18,7,2,0,
+41,8,1,0,77,6,4,128,32,10,5,9,59,11,9,0,64,11,10,4,79,6,252,127,88,6,8,128,
+18,6,2,0,18,7,1,0,41,8,255,255,77,6,4,128,32,10,5,9,59,11,9,0,64,11,10,4,79,
+6,252,127,76,4,2,0,0
+#else
+0,1,2,0,0,1,2,24,1,0,0,76,1,2,0,241,135,158,166,3,220,203,178,130,4,0,1,2,0,
+0,1,2,24,1,0,0,76,1,2,0,243,244,148,165,20,198,190,199,252,3,0,1,2,0,0,0,3,
+16,0,5,0,21,1,0,0,76,1,2,0,0,2,9,0,0,0,15,16,0,12,0,16,1,9,0,41,2,1,0,21,3,
+0,0,41,4,1,0,77,2,8,128,18,6,1,0,18,7,5,0,59,8,5,0,66,6,3,2,10,6,0,0,88,7,1,
+128,76,6,2,0,79,2,248,127,75,0,1,0,0,2,10,0,0,0,16,16,0,12,0,16,1,9,0,43,2,
+0,0,18,3,0,0,41,4,0,0,88,5,7,128,18,7,1,0,18,8,5,0,18,9,6,0,66,7,3,2,10,7,0,
+0,88,8,1,128,76,7,2,0,70,5,3,3,82,5,247,127,75,0,1,0,0,1,2,0,0,0,3,16,0,12,
+0,21,1,0,0,76,1,2,0,0,2,10,0,0,2,30,16,0,12,0,21,2,0,0,11,1,0,0,88,3,7,128,
+8,2,0,0,88,3,23,128,59,3,2,0,43,4,0,0,64,4,2,0,76,3,2,0,88,3,18,128,16,1,14,
+0,41,3,1,0,3,3,1,0,88,3,14,128,3,1,2,0,88,3,12,128,59,3,1,0,22,4,1,1,18,5,2,
+0,41,6,1,0,77,4,4,128,23,8,1,7,59,9,7,0,64,9,8,0,79,4,252,127,43,4,0,0,64,4,
+2,0,76,3,2,0,75,0,1,0,0,2,0,5,12,0,0,0,35,16,0,12,0,16,1,14,0,16,2,14,0,16,
+3,14,0,11,4,0,0,88,5,1,128,18,4,0,0,16,4,12,0,3,1,2,0,88,5,24,128,33,5,1,3,
+0,2,3,0,88,6,4,128,2,3,1,0,88,6,2,128,4,4,0,0,88,6,9,128,18,6,1,0,18,7,2,0,
+41,8,1,0,77,6,4,128,32,10,5,9,59,11,9,0,64,11,10,4,79,6,252,127,88,6,8,128,
+18,6,2,0,18,7,1,0,41,8,255,255,77,6,4,128,32,10,5,9,59,11,9,0,64,11,10,4,79,
+6,252,127,76,4,2,0,0
+#endif
+};
+
+static const struct { const char *name; int ofs; } libbc_map[] = {
+{"math_deg",0},
+{"math_rad",25},
+{"string_len",50},
+{"table_foreachi",69},
+{"table_foreach",136},
+{"table_getn",207},
+{"table_remove",226},
+{"table_move",355},
+{NULL,502}
+};
+
diff --git a/src/host/buildvm_peobj.c b/src/host/buildvm_peobj.c
index aab00d68..2eb2bb7b 100644
--- a/src/host/buildvm_peobj.c
+++ b/src/host/buildvm_peobj.c
@@ -109,6 +109,8 @@ enum {
 #if LJ_TARGET_X64
   PEOBJ_SECT_PDATA,
   PEOBJ_SECT_XDATA,
+#elif LJ_TARGET_X86
+  PEOBJ_SECT_SXDATA,
 #endif
   PEOBJ_SECT_RDATA_Z,
   PEOBJ_NSECTIONS
@@ -208,6 +210,13 @@ void emit_peobj(BuildCtx *ctx)
   sofs += (pesect[PEOBJ_SECT_XDATA].nreloc = 1) * PEOBJ_RELOC_SIZE;
   /* Flags: 40 = read, 30 = align4, 40 = initialized data. */
   pesect[PEOBJ_SECT_XDATA].flags = 0x40300040;
+#elif LJ_TARGET_X86
+  memcpy(pesect[PEOBJ_SECT_SXDATA].name, ".sxdata", sizeof(".sxdata")-1);
+  pesect[PEOBJ_SECT_SXDATA].ofs = sofs;
+  sofs += (pesect[PEOBJ_SECT_SXDATA].size = 4);
+  pesect[PEOBJ_SECT_SXDATA].relocofs = sofs;
+  /* Flags: 40 = read, 30 = align4, 02 = lnk_info, 40 = initialized data. */
+  pesect[PEOBJ_SECT_SXDATA].flags = 0x40300240;
 #endif
 
   memcpy(pesect[PEOBJ_SECT_RDATA_Z].name, ".rdata$Z", sizeof(".rdata$Z")-1);
@@ -232,7 +241,7 @@ void emit_peobj(BuildCtx *ctx)
   nrsym = ctx->nrelocsym;
   pehdr.nsyms = 1+PEOBJ_NSECTIONS*2 + 1+ctx->nsym + nrsym;
 #if LJ_TARGET_X64
-  pehdr.nsyms += 1;  /* Symbol for lj_err_unwind_win64. */
+  pehdr.nsyms += 1;  /* Symbol for lj_err_unwind_win. */
 #endif
 
   /* Write PE object header and all sections. */
@@ -312,6 +321,19 @@ void emit_peobj(BuildCtx *ctx)
     reloc.type = PEOBJ_RELOC_ADDR32NB;
     owrite(ctx, &reloc, PEOBJ_RELOC_SIZE);
   }
+#elif LJ_TARGET_X86
+  /* Write .sxdata section. */
+  for (i = 0; i < nrsym; i++) {
+    if (!strcmp(ctx->relocsym[i], "_lj_err_unwind_win")) {
+      uint32_t symidx = 1+2+i;
+      owrite(ctx, &symidx, 4);
+      break;
+    }
+  }
+  if (i == nrsym) {
+    fprintf(stderr, "Error: extern lj_err_unwind_win not used\n");
+    exit(1);
+  }
 #endif
 
   /* Write .rdata$Z section. */
@@ -333,8 +355,10 @@ void emit_peobj(BuildCtx *ctx)
 #if LJ_TARGET_X64
     emit_peobj_sym_sect(ctx, pesect, PEOBJ_SECT_PDATA);
     emit_peobj_sym_sect(ctx, pesect, PEOBJ_SECT_XDATA);
-    emit_peobj_sym(ctx, "lj_err_unwind_win64", 0,
+    emit_peobj_sym(ctx, "lj_err_unwind_win", 0,
                    PEOBJ_SECT_UNDEF, PEOBJ_TYPE_FUNC, PEOBJ_SCL_EXTERN);
+#elif LJ_TARGET_X86
+    emit_peobj_sym_sect(ctx, pesect, PEOBJ_SECT_SXDATA);
 #endif
 
     emit_peobj_sym(ctx, ctx->beginsym, 0,
diff --git a/src/host/genlibbc.lua b/src/host/genlibbc.lua
new file mode 100644
index 00000000..6f5a05cc
--- /dev/null
+++ b/src/host/genlibbc.lua
@@ -0,0 +1,197 @@
+----------------------------------------------------------------------------
+-- 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) 2005-2017 Mike Pall. All rights reserved.
+-- Released under the MIT license. See Copyright Notice in luajit.h
+----------------------------------------------------------------------------
+
+local ffi = require("ffi")
+local bit = require("bit")
+local vmdef = require("jit.vmdef")
+local bcnames = vmdef.bcnames
+
+local format = string.format
+
+local isbe = (string.byte(string.dump(function() end), 5) % 2 == 1)
+
+local function usage(arg)
+  io.stderr:write("Usage: ", arg and arg[0] or "genlibbc",
+                  " [-o buildvm_libbc.h] lib_*.c\n")
+  os.exit(1)
+end
+
+local function parse_arg(arg)
+  local outfile = "-"
+  if not (arg and arg[1]) then
+    usage(arg)
+  end
+  if arg[1] == "-o" then
+    outfile = arg[2]
+    if not outfile then usage(arg) end
+    table.remove(arg, 1)
+    table.remove(arg, 1)
+  end
+  return outfile
+end
+
+local function read_files(names)
+  local src = ""
+  for _,name in ipairs(names) do
+    local fp = assert(io.open(name))
+    src = src .. fp:read("*a")
+    fp:close()
+  end
+  return src
+end
+
+local function transform_lua(code)
+  local fixup = {}
+  local n = -30000
+  code = string.gsub(code, "CHECK_(%w*)%((.-)%)", function(tp, var)
+    n = n + 1
+    fixup[n] = { "CHECK", tp }
+    return format("%s=%d", var, n)
+  end)
+  code = string.gsub(code, "PAIRS%((.-)%)", function(var)
+    fixup.PAIRS = true
+    return format("nil, %s, 0", var)
+  end)
+  return "return "..code, fixup
+end
+
+local function read_uleb128(p)
+  local v = p[0]; p = p + 1
+  if v >= 128 then
+    local sh = 7; v = v - 128
+    repeat
+      local r = p[0]
+      v = v + bit.lshift(bit.band(r, 127), sh)
+      sh = sh + 7
+      p = p + 1
+    until r < 128
+  end
+  return p, v
+end
+
+-- ORDER LJ_T
+local name2itype = {
+  str = 5, func = 9, tab = 12, int = 14, num = 15
+}
+
+local BC = {}
+for i=0,#bcnames/6-1 do
+  BC[string.gsub(string.sub(bcnames, i*6+1, i*6+6), " ", "")] = i
+end
+local xop, xra = isbe and 3 or 0, isbe and 2 or 1
+local xrc, xrb = isbe and 1 or 2, isbe and 0 or 3
+
+local function fixup_dump(dump, fixup)
+  local buf = ffi.new("uint8_t[?]", #dump+1, dump)
+  local p = buf+5
+  local n, sizebc
+  p, n = read_uleb128(p)
+  local start = p
+  p = p + 4
+  p = read_uleb128(p)
+  p = read_uleb128(p)
+  p, sizebc = read_uleb128(p)
+  local rawtab = {}
+  for i=0,sizebc-1 do
+    local op = p[xop]
+    if op == BC.KSHORT then
+      local rd = p[xrc] + 256*p[xrb]
+      rd = bit.arshift(bit.lshift(rd, 16), 16)
+      local f = fixup[rd]
+      if f then
+        if f[1] == "CHECK" then
+          local tp = f[2]
+          if tp == "tab" then rawtab[p[xra]] = true end
+          p[xop] = tp == "num" and BC.ISNUM or BC.ISTYPE
+          p[xrb] = 0
+          p[xrc] = name2itype[tp]
+        else
+          error("unhandled fixup type: "..f[1])
+        end
+      end
+    elseif op == BC.TGETV then
+      if rawtab[p[xrb]] then
+        p[xop] = BC.TGETR
+      end
+    elseif op == BC.TSETV then
+      if rawtab[p[xrb]] then
+        p[xop] = BC.TSETR
+      end
+    elseif op == BC.ITERC then
+      if fixup.PAIRS then
+        p[xop] = BC.ITERN
+      end
+    end
+    p = p + 4
+  end
+  return ffi.string(start, n)
+end
+
+local function find_defs(src)
+  local defs = {}
+  for name, code in string.gmatch(src, "LJLIB_LUA%(([^)]*)%)%s*/%*(.-)%*/") do
+    local env = {}
+    local tcode, fixup = transform_lua(code)
+    local func = assert(load(tcode, "", nil, env))()
+    defs[name] = fixup_dump(string.dump(func, true), fixup)
+    defs[#defs+1] = name
+  end
+  return defs
+end
+
+local function gen_header(defs)
+  local t = {}
+  local function w(x) t[#t+1] = x end
+  w("/* This is a generated file. DO NOT EDIT! */\n\n")
+  w("static const int libbc_endian = ") w(isbe and 1 or 0) w(";\n\n")
+  local s = ""
+  for _,name in ipairs(defs) do
+    s = s .. defs[name]
+  end
+  w("static const uint8_t libbc_code[] = {\n")
+  local n = 0
+  for i=1,#s do
+    local x = string.byte(s, i)
+    w(x); w(",")
+    n = n + (x < 10 and 2 or (x < 100 and 3 or 4))
+    if n >= 75 then n = 0; w("\n") end
+  end
+  w("0\n};\n\n")
+  w("static const struct { const char *name; int ofs; } libbc_map[] = {\n")
+  local m = 0
+  for _,name in ipairs(defs) do
+    w('{"'); w(name); w('",'); w(m) w('},\n')
+    m = m + #defs[name]
+  end
+  w("{NULL,"); w(m); w("}\n};\n\n")
+  return table.concat(t)
+end
+
+local function write_file(name, data)
+  if name == "-" then
+    assert(io.write(data))
+    assert(io.flush())
+  else
+    local fp = io.open(name)
+    if fp then
+      local old = fp:read("*a")
+      fp:close()
+      if data == old then return end
+    end
+    fp = assert(io.open(name, "w"))
+    assert(fp:write(data))
+    assert(fp:close())
+  end
+end
+
+local outfile = parse_arg(arg)
+local src = read_files(arg)
+local defs = find_defs(src)
+local hdr = gen_header(defs)
+write_file(outfile, hdr)
+