Merge commit 'f159ed96b9a02dd99ea1c0c2f4a71f506de0697c' into dev

src/deps/src/luajit/.travis.yml

@@ -10,6 +10,7 @@ arch:
   - amd64
   - arm64
   - s390x
+  - ppc64le
 
 language: c
 

src/deps/src/luajit/doc/ext_buffer.html

@@ -85,7 +85,7 @@ <h1>String Buffer Library</h1>
 </p>
 <p>
 The string buffer library also includes a high-performance
-<a href="serialize">serializer</a> for Lua objects.
+<a href="#serialize">serializer</a> for Lua objects.
 </p>
 
 <h2 id="use">Using the String Buffer Library</h2>
@@ -588,9 +588,9 @@ <h3 id="serialize_format">Serialization Format Specification</h3>
 tab       → 0x08                                   // Empty table
           | 0x09 h.U h*{object object}          // Key/value hash
           | 0x0a a.U a*object                    // 0-based array
-          | 0x0b a.U a*object h.U h*{object object}      // Mixed
+          | 0x0b a.U h.U a*object h*{object object}      // Mixed
           | 0x0c a.U (a-1)*object                // 1-based array
-          | 0x0d a.U (a-1)*object h.U h*{object object}  // Mixed
+          | 0x0d a.U h.U (a-1)*object h*{object object}  // Mixed
 tab_mt    → 0x0e (index-1).U tab          // Metatable dict entry
 
 int64     → 0x10 int.L                             // FFI int64_t

src/deps/src/luajit/doc/ext_ffi_semantics.html

@@ -440,6 +440,19 @@ <h3 id="convert_vararg">Conversions for vararg C&nbsp;function arguments</h3>
 conversion rule applies. A vararg C&nbsp;function expecting an integer
 will see a garbled or uninitialized value.
 </p>
+<p>
+Note: this is the only place where creating a boxed scalar number type is
+actually useful. <b>Never use <tt>ffi.new("int")</tt>, <tt>ffi.new("float")</tt>
+etc. anywhere else!</b>
+</p>
+<p style="font-size: 8pt;">
+Ditto for <tt>ffi.cast()</tt>. Explicitly boxing scalars <b>does not</b>
+improve performance or force <tt>int</tt> or <tt>float</tt> arithmetic! It
+just adds costly boxing, unboxing and conversions steps. And it may lead
+to surprise results, because
+<a href="#cdata_arith">cdata arithmetic on scalar numbers</a>
+is always performed on 64 bit integers.
+</p>
 
 <h2 id="init">Initializers</h2>
 <p>

src/deps/src/luajit/src/jit/bcsave.lua

@@ -441,24 +441,12 @@ typedef struct
 {
   mach_header; uint32_t reserved;
 } mach_header_64;
-typedef struct {
-  uint32_t cmd, cmdsize;
-  char segname[16];
-  uint32_t vmaddr, vmsize, fileoff, filesize;
-  uint32_t maxprot, initprot, nsects, flags;
-} mach_segment_command;
 typedef struct {
   uint32_t cmd, cmdsize;
   char segname[16];
   uint64_t vmaddr, vmsize, fileoff, filesize;
   uint32_t maxprot, initprot, nsects, flags;
 } mach_segment_command_64;
-typedef struct {
-  char sectname[16], segname[16];
-  uint32_t addr, size;
-  uint32_t offset, align, reloff, nreloc, flags;
-  uint32_t reserved1, reserved2;
-} mach_section;
 typedef struct {
   char sectname[16], segname[16];
   uint64_t addr, size;
@@ -468,133 +456,58 @@ typedef struct {
 typedef struct {
   uint32_t cmd, cmdsize, symoff, nsyms, stroff, strsize;
 } mach_symtab_command;
-typedef struct {
-  int32_t strx;
-  uint8_t type, sect;
-  int16_t desc;
-  uint32_t value;
-} mach_nlist;
 typedef struct {
   int32_t strx;
   uint8_t type, sect;
   uint16_t desc;
   uint64_t value;
 } mach_nlist_64;
-typedef struct
-{
-  int32_t magic, nfat_arch;
-} mach_fat_header;
-typedef struct
-{
-  int32_t cputype, cpusubtype, offset, size, align;
-} mach_fat_arch;
-typedef struct {
-  struct {
-    mach_header hdr;
-    mach_segment_command seg;
-    mach_section sec;
-    mach_symtab_command sym;
-  } arch[1];
-  mach_nlist sym_entry;
-  uint8_t space[4096];
-} mach_obj;
 typedef struct {
-  struct {
-    mach_header_64 hdr;
-    mach_segment_command_64 seg;
-    mach_section_64 sec;
-    mach_symtab_command sym;
-  } arch[1];
+  mach_header_64 hdr;
+  mach_segment_command_64 seg;
+  mach_section_64 sec;
+  mach_symtab_command sym;
   mach_nlist_64 sym_entry;
   uint8_t space[4096];
 } mach_obj_64;
-typedef struct {
-  mach_fat_header fat;
-  mach_fat_arch fat_arch[2];
-  struct {
-    mach_header hdr;
-    mach_segment_command seg;
-    mach_section sec;
-    mach_symtab_command sym;
-  } arch[2];
-  mach_nlist sym_entry;
-  uint8_t space[4096];
-} mach_fat_obj;
-typedef struct {
-  mach_fat_header fat;
-  mach_fat_arch fat_arch[2];
-  struct {
-    mach_header_64 hdr;
-    mach_segment_command_64 seg;
-    mach_section_64 sec;
-    mach_symtab_command sym;
-  } arch[2];
-  mach_nlist_64 sym_entry;
-  uint8_t space[4096];
-} mach_fat_obj_64;
 ]]
   local symname = '_'..LJBC_PREFIX..ctx.modname
-  local isfat, is64, align, mobj = false, false, 4, "mach_obj"
-  if ctx.arch == "x64" then
-    is64, align, mobj = true, 8, "mach_obj_64"
-  elseif ctx.arch == "arm" then
-    isfat, mobj = true, "mach_fat_obj"
-  elseif ctx.arch == "arm64" then
-    is64, align, isfat, mobj = true, 8, true, "mach_fat_obj_64"
-  else
-    check(ctx.arch == "x86", "unsupported architecture for OSX")
+  local cputype, cpusubtype = 0x01000007, 3
+  if ctx.arch ~= "x64" then
+    check(ctx.arch == "arm64", "unsupported architecture for OSX")
+    cputype, cpusubtype = 0x0100000c, 0
   end
   local function aligned(v, a) return bit.band(v+a-1, -a) end
-  local be32 = bit.bswap -- Mach-O FAT is BE, supported archs are LE.
 
   -- Create Mach-O object and fill in header.
-  local o = ffi.new(mobj)
-  local mach_size = aligned(ffi.offsetof(o, "space")+#symname+2, align)
-  local cputype = ({ x86={7}, x64={0x01000007}, arm={7,12}, arm64={0x01000007,0x0100000c} })[ctx.arch]
-  local cpusubtype = ({ x86={3}, x64={3}, arm={3,9}, arm64={3,0} })[ctx.arch]
-  if isfat then
-    o.fat.magic = be32(0xcafebabe)
-    o.fat.nfat_arch = be32(#cpusubtype)
-  end
+  local o = ffi.new("mach_obj_64")
+  local mach_size = aligned(ffi.offsetof(o, "space")+#symname+2, 8)
 
   -- Fill in sections and symbols.
-  for i=0,#cpusubtype-1 do
-    local ofs = 0
-    if isfat then
-      local a = o.fat_arch[i]
-      a.cputype = be32(cputype[i+1])
-      a.cpusubtype = be32(cpusubtype[i+1])
-      -- Subsequent slices overlap each other to share data.
-      ofs = ffi.offsetof(o, "arch") + i*ffi.sizeof(o.arch[0])
-      a.offset = be32(ofs)
-      a.size = be32(mach_size-ofs+#s)
-    end
-    local a = o.arch[i]
-    a.hdr.magic = is64 and 0xfeedfacf or 0xfeedface
-    a.hdr.cputype = cputype[i+1]
-    a.hdr.cpusubtype = cpusubtype[i+1]
-    a.hdr.filetype = 1
-    a.hdr.ncmds = 2
-    a.hdr.sizeofcmds = ffi.sizeof(a.seg)+ffi.sizeof(a.sec)+ffi.sizeof(a.sym)
-    a.seg.cmd = is64 and 0x19 or 0x1
-    a.seg.cmdsize = ffi.sizeof(a.seg)+ffi.sizeof(a.sec)
-    a.seg.vmsize = #s
-    a.seg.fileoff = mach_size-ofs
-    a.seg.filesize = #s
-    a.seg.maxprot = 1
-    a.seg.initprot = 1
-    a.seg.nsects = 1
-    ffi.copy(a.sec.sectname, "__data")
-    ffi.copy(a.sec.segname, "__DATA")
-    a.sec.size = #s
-    a.sec.offset = mach_size-ofs
-    a.sym.cmd = 2
-    a.sym.cmdsize = ffi.sizeof(a.sym)
-    a.sym.symoff = ffi.offsetof(o, "sym_entry")-ofs
-    a.sym.nsyms = 1
-    a.sym.stroff = ffi.offsetof(o, "sym_entry")+ffi.sizeof(o.sym_entry)-ofs
-    a.sym.strsize = aligned(#symname+2, align)
-  end
+  o.hdr.magic = 0xfeedfacf
+  o.hdr.cputype = cputype
+  o.hdr.cpusubtype = cpusubtype
+  o.hdr.filetype = 1
+  o.hdr.ncmds = 2
+  o.hdr.sizeofcmds = ffi.sizeof(o.seg)+ffi.sizeof(o.sec)+ffi.sizeof(o.sym)
+  o.seg.cmd = 0x19
+  o.seg.cmdsize = ffi.sizeof(o.seg)+ffi.sizeof(o.sec)
+  o.seg.vmsize = #s
+  o.seg.fileoff = mach_size
+  o.seg.filesize = #s
+  o.seg.maxprot = 1
+  o.seg.initprot = 1
+  o.seg.nsects = 1
+  ffi.copy(o.sec.sectname, "__data")
+  ffi.copy(o.sec.segname, "__DATA")
+  o.sec.size = #s
+  o.sec.offset = mach_size
+  o.sym.cmd = 2
+  o.sym.cmdsize = ffi.sizeof(o.sym)
+  o.sym.symoff = ffi.offsetof(o, "sym_entry")
+  o.sym.nsyms = 1
+  o.sym.stroff = ffi.offsetof(o, "sym_entry")+ffi.sizeof(o.sym_entry)
+  o.sym.strsize = aligned(#symname+2, 8)
   o.sym_entry.type = 0xf
   o.sym_entry.sect = 1
   o.sym_entry.strx = 1

src/deps/src/luajit/src/jit/dump.lua

@@ -552,7 +552,12 @@ local recdepth = 0
 local function fmterr(err, info)
   if type(err) == "number" then
     if type(info) == "function" then info = fmtfunc(info) end
-    err = format(vmdef.traceerr[err], info)
+    local fmt = vmdef.traceerr[err]
+    if fmt == "NYI: bytecode %s" then
+      local oidx = 6 * info
+      info = sub(vmdef.bcnames, oidx+1, oidx+6)
+    end
+    err = format(fmt, info)
   end
   return err
 end

src/deps/src/luajit/src/jit/v.lua

@@ -62,7 +62,7 @@ local jit = require("jit")
 local jutil = require("jit.util")
 local vmdef = require("jit.vmdef")
 local funcinfo, traceinfo = jutil.funcinfo, jutil.traceinfo
-local type, format = type, string.format
+local type, sub, format = type, string.sub, string.format
 local stdout, stderr = io.stdout, io.stderr
 
 -- Active flag and output file handle.
@@ -89,7 +89,12 @@ end
 local function fmterr(err, info)
   if type(err) == "number" then
     if type(info) == "function" then info = fmtfunc(info) end
-    err = format(vmdef.traceerr[err], info)
+    local fmt = vmdef.traceerr[err]
+    if fmt == "NYI: bytecode %s" then
+      local oidx = 6 * info
+      info = sub(vmdef.bcnames, oidx+1, oidx+6)
+    end
+    err = format(fmt, info)
   end
   return err
 end

src/deps/src/luajit/src/lib_ffi.c

@@ -513,7 +513,7 @@ LJLIB_CF(ffi_new)	LJLIB_REC(.)
     /* Handle ctype __gc metamethod. Use the fast lookup here. */
     cTValue *tv = lj_tab_getinth(cts->miscmap, -(int32_t)id);
     if (tv && tvistab(tv) && (tv = lj_meta_fast(L, tabV(tv), MM_gc))) {
-      GCtab *t = cts->finalizer;
+      GCtab *t = tabref(G(L)->gcroot[GCROOT_FFI_FIN]);
       if (gcref(t->metatable)) {
 	/* Add to finalizer table, if still enabled. */
 	copyTV(L, lj_tab_set(L, t, o-1), tv);
@@ -765,7 +765,7 @@ LJLIB_CF(ffi_abi)	LJLIB_REC(.)
   return 1;
 }
 
-LJLIB_PUSH(top-8) LJLIB_SET(!)  /* Store reference to miscmap table. */
+LJLIB_PUSH(top-7) LJLIB_SET(!)  /* Store reference to miscmap table. */
 
 LJLIB_CF(ffi_metatype)
 {
@@ -791,8 +791,6 @@ LJLIB_CF(ffi_metatype)
   return 1;
 }
 
-LJLIB_PUSH(top-7) LJLIB_SET(!)  /* Store reference to finalizer table. */
-
 LJLIB_CF(ffi_gc)	LJLIB_REC(.)
 {
   GCcdata *cd = ffi_checkcdata(L, 1);
@@ -825,19 +823,6 @@ LJLIB_PUSH(top-2) LJLIB_SET(arch)
 
 /* ------------------------------------------------------------------------ */
 
-/* Create special weak-keyed finalizer table. */
-static GCtab *ffi_finalizer(lua_State *L)
-{
-  /* NOBARRIER: The table is new (marked white). */
-  GCtab *t = lj_tab_new(L, 0, 1);
-  settabV(L, L->top++, t);
-  setgcref(t->metatable, obj2gco(t));
-  setstrV(L, lj_tab_setstr(L, t, lj_str_newlit(L, "__mode")),
-	  lj_str_newlit(L, "k"));
-  t->nomm = (uint8_t)(~(1u<<MM_mode));
-  return t;
-}
-
 /* Register FFI module as loaded. */
 static void ffi_register_module(lua_State *L)
 {
@@ -853,7 +838,6 @@ LUALIB_API int luaopen_ffi(lua_State *L)
 {
   CTState *cts = lj_ctype_init(L);
   settabV(L, L->top++, (cts->miscmap = lj_tab_new(L, 0, 1)));
-  cts->finalizer = ffi_finalizer(L);
   LJ_LIB_REG(L, NULL, ffi_meta);
   /* NOBARRIER: basemt is a GC root. */
   setgcref(basemt_it(G(L), LJ_TCDATA), obj2gco(tabV(L->top-1)));

src/deps/src/luajit/src/lj_alloc.c

@@ -1057,7 +1057,7 @@ static size_t release_unused_segments(mstate m)
       mchunkptr p = align_as_chunk(base);
       size_t psize = chunksize(p);
       /* Can unmap if first chunk holds entire segment and not pinned */
-      if (!cinuse(p) && (char *)p + psize >= base + size - TOP_FOOT_SIZE) {
+      if (!cinuse(p) && (char *)p + psize == (char *)mem2chunk(sp)) {
 	tchunkptr tp = (tchunkptr)p;
 	if (p == m->dv) {
 	  m->dv = 0;

src/deps/src/luajit/src/lj_arch.h

@@ -128,7 +128,7 @@
 #define LJ_TARGET_POSIX		(LUAJIT_OS > LUAJIT_OS_WINDOWS)
 #define LJ_TARGET_DLOPEN	LJ_TARGET_POSIX
 
-#if TARGET_OS_IPHONE
+#if defined(TARGET_OS_IPHONE) && TARGET_OS_IPHONE
 #define LJ_TARGET_IOS		1
 #else
 #define LJ_TARGET_IOS		0

src/deps/src/luajit/src/lj_asm_arm64.h

@@ -816,7 +816,7 @@ static void asm_href(ASMState *as, IRIns *ir, IROp merge)
   int destused = ra_used(ir);
   Reg dest = ra_dest(as, ir, allow);
   Reg tab = ra_alloc1(as, ir->op1, rset_clear(allow, dest));
-  Reg tmp = RID_TMP, type = RID_NONE, key, tkey;
+  Reg tmp = RID_TMP, type = RID_NONE, key = RID_NONE, tkey;
   IRRef refkey = ir->op2;
   IRIns *irkey = IR(refkey);
   int isk = irref_isk(refkey);

src/deps/src/luajit/src/lj_cdata.c

@@ -86,7 +86,7 @@ void LJ_FASTCALL lj_cdata_free(global_State *g, GCcdata *cd)
 
 void lj_cdata_setfin(lua_State *L, GCcdata *cd, GCobj *obj, uint32_t it)
 {
-  GCtab *t = ctype_ctsG(G(L))->finalizer;
+  GCtab *t = tabref(G(L)->gcroot[GCROOT_FFI_FIN]);
   if (gcref(t->metatable)) {
     /* Add cdata to finalizer table, if still enabled. */
     TValue *tv, tmp;

src/deps/src/luajit/src/lj_ctype.c

@@ -643,6 +643,18 @@ CTState *lj_ctype_init(lua_State *L)
   return cts;
 }
 
+/* Create special weak-keyed finalizer table. */
+void lj_ctype_initfin(lua_State *L)
+{
+  /* NOBARRIER: The table is new (marked white). */
+  GCtab *t = lj_tab_new(L, 0, 1);
+  setgcref(t->metatable, obj2gco(t));
+  setstrV(L, lj_tab_setstr(L, t, lj_str_newlit(L, "__mode")),
+	  lj_str_newlit(L, "k"));
+  t->nomm = (uint8_t)(~(1u<<MM_mode));
+  setgcref(G(L)->gcroot[GCROOT_FFI_FIN], obj2gco(t));
+}
+
 /* Free C type table and state. */
 void lj_ctype_freestate(global_State *g)
 {