Update embedded LUA to 5.4.4

Signed-off-by: DL6ER <dl6er@dl6er.de>

src/lua/lcode.c

@@ -10,6 +10,7 @@
 #include "lprefix.h"
 
 
+#include <float.h>
 #include <limits.h>
 #include <math.h>
 #include <stdlib.h>
@@ -314,15 +315,6 @@ void luaK_patchtohere (FuncState *fs, int list) {
 }
 
 
-/*
-** MAXimum number of successive Instructions WiTHout ABSolute line
-** information.
-*/
-#if !defined(MAXIWTHABS)
-#define MAXIWTHABS	120
-#endif
-
-
 /* limit for difference between lines in relative line info. */
 #define LIMLINEDIFF	0x80
 
@@ -337,13 +329,13 @@ void luaK_patchtohere (FuncState *fs, int list) {
 static void savelineinfo (FuncState *fs, Proto *f, int line) {
   int linedif = line - fs->previousline;
   int pc = fs->pc - 1;  /* last instruction coded */
-  if (abs(linedif) >= LIMLINEDIFF || fs->iwthabs++ > MAXIWTHABS) {
+  if (abs(linedif) >= LIMLINEDIFF || fs->iwthabs++ >= MAXIWTHABS) {
     luaM_growvector(fs->ls->L, f->abslineinfo, fs->nabslineinfo,
                     f->sizeabslineinfo, AbsLineInfo, MAX_INT, "lines");
     f->abslineinfo[fs->nabslineinfo].pc = pc;
     f->abslineinfo[fs->nabslineinfo++].line = line;
     linedif = ABSLINEINFO;  /* signal that there is absolute information */
-    fs->iwthabs = 0;  /* restart counter */
+    fs->iwthabs = 1;  /* restart counter */
   }
   luaM_growvector(fs->ls->L, f->lineinfo, pc, f->sizelineinfo, ls_byte,
                   MAX_INT, "opcodes");
@@ -545,11 +537,14 @@ static void freeexps (FuncState *fs, expdesc *e1, expdesc *e2) {
 ** and try to reuse constants. Because some values should not be used
 ** as keys (nil cannot be a key, integer keys can collapse with float
 ** keys), the caller must provide a useful 'key' for indexing the cache.
+** Note that all functions share the same table, so entering or exiting
+** a function can make some indices wrong.
 */
 static int addk (FuncState *fs, TValue *key, TValue *v) {
+  TValue val;
   lua_State *L = fs->ls->L;
   Proto *f = fs->f;
-  TValue *idx = luaH_set(L, fs->ls->h, key);  /* index scanner table */
+  const TValue *idx = luaH_get(fs->ls->h, key);  /* query scanner table */
   int k, oldsize;
   if (ttisinteger(idx)) {  /* is there an index there? */
     k = cast_int(ivalue(idx));
@@ -563,7 +558,8 @@ static int addk (FuncState *fs, TValue *key, TValue *v) {
   k = fs->nk;
   /* numerical value does not need GC barrier;
      table has no metatable, so it does not need to invalidate cache */
-  setivalue(idx, k);
+  setivalue(&val, k);
+  luaH_finishset(L, fs->ls->h, key, idx, &val);
   luaM_growvector(L, f->k, k, f->sizek, TValue, MAXARG_Ax, "constants");
   while (oldsize < f->sizek) setnilvalue(&f->k[oldsize++]);
   setobj(L, &f->k[k], v);
@@ -585,24 +581,41 @@ static int stringK (FuncState *fs, TString *s) {
 
 /*
 ** Add an integer to list of constants and return its index.
-** Integers use userdata as keys to avoid collision with floats with
-** same value; conversion to 'void*' is used only for hashing, so there
-** are no "precision" problems.
 */
 static int luaK_intK (FuncState *fs, lua_Integer n) {
-  TValue k, o;
-  setpvalue(&k, cast_voidp(cast_sizet(n)));
+  TValue o;
   setivalue(&o, n);
-  return addk(fs, &k, &o);
+  return addk(fs, &o, &o);  /* use integer itself as key */
 }
 
 /*
-** Add a float to list of constants and return its index.
+** Add a float to list of constants and return its index. Floats
+** with integral values need a different key, to avoid collision
+** with actual integers. To that, we add to the number its smaller
+** power-of-two fraction that is still significant in its scale.
+** For doubles, that would be 1/2^52.
+** (This method is not bulletproof: there may be another float
+** with that value, and for floats larger than 2^53 the result is
+** still an integer. At worst, this only wastes an entry with
+** a duplicate.)
 */
 static int luaK_numberK (FuncState *fs, lua_Number r) {
   TValue o;
+  lua_Integer ik;
   setfltvalue(&o, r);
-  return addk(fs, &o, &o);  /* use number itself as key */
+  if (!luaV_flttointeger(r, &ik, F2Ieq))  /* not an integral value? */
+    return addk(fs, &o, &o);  /* use number itself as key */
+  else {  /* must build an alternative key */
+    const int nbm = l_floatatt(MANT_DIG);
+    const lua_Number q = l_mathop(ldexp)(l_mathop(1.0), -nbm + 1);
+    const lua_Number k = (ik == 0) ? q : r + r*q;  /* new key */
+    TValue kv;
+    setfltvalue(&kv, k);
+    /* result is not an integral value, unless value is too large */
+    lua_assert(!luaV_flttointeger(k, &ik, F2Ieq) ||
+                l_mathop(fabs)(r) >= l_mathop(1e6));
+    return addk(fs, &kv, &o);
+  }
 }
 
 
@@ -753,7 +766,7 @@ void luaK_setoneret (FuncState *fs, expdesc *e) {
 
 
 /*
-** Ensure that expression 'e' is not a variable (nor a constant).
+** Ensure that expression 'e' is not a variable (nor a <const>).
 ** (Expression still may have jump lists.)
 */
 void luaK_dischargevars (FuncState *fs, expdesc *e) {
@@ -763,7 +776,7 @@ void luaK_dischargevars (FuncState *fs, expdesc *e) {
       break;
     }
     case VLOCAL: {  /* already in a register */
-      e->u.info = e->u.var.sidx;
+      e->u.info = e->u.var.ridx;
       e->k = VNONRELOC;  /* becomes a non-relocatable value */
       break;
     }
@@ -805,8 +818,8 @@ void luaK_dischargevars (FuncState *fs, expdesc *e) {
 
 
 /*
-** Ensures expression value is in register 'reg' (and therefore
-** 'e' will become a non-relocatable expression).
+** Ensure expression value is in register 'reg', making 'e' a
+** non-relocatable expression.
 ** (Expression still may have jump lists.)
 */
 static void discharge2reg (FuncState *fs, expdesc *e, int reg) {
@@ -860,7 +873,8 @@ static void discharge2reg (FuncState *fs, expdesc *e, int reg) {
 
 
 /*
-** Ensures expression value is in any register.
+** Ensure expression value is in a register, making 'e' a
+** non-relocatable expression.
 ** (Expression still may have jump lists.)
 */
 static void discharge2anyreg (FuncState *fs, expdesc *e) {
@@ -946,8 +960,11 @@ int luaK_exp2anyreg (FuncState *fs, expdesc *e) {
       exp2reg(fs, e, e->u.info);  /* put final result in it */
       return e->u.info;
     }
+    /* else expression has jumps and cannot change its register
+       to hold the jump values, because it is a local variable.
+       Go through to the default case. */
   }
-  luaK_exp2nextreg(fs, e);  /* otherwise, use next available register */
+  luaK_exp2nextreg(fs, e);  /* default: use next available register */
   return e->u.info;
 }
 
@@ -1032,7 +1049,7 @@ void luaK_storevar (FuncState *fs, expdesc *var, expdesc *ex) {
   switch (var->k) {
     case VLOCAL: {
       freeexp(fs, ex);
-      exp2reg(fs, ex, var->u.var.sidx);  /* compute 'ex' into proper place */
+      exp2reg(fs, ex, var->u.var.ridx);  /* compute 'ex' into proper place */
       return;
     }
     case VUPVAL: {
@@ -1272,7 +1289,7 @@ void luaK_indexed (FuncState *fs, expdesc *t, expdesc *k) {
   }
   else {
     /* register index of the table */
-    t->u.ind.t = (t->k == VLOCAL) ? t->u.var.sidx: t->u.info;
+    t->u.ind.t = (t->k == VLOCAL) ? t->u.var.ridx: t->u.info;
     if (isKstr(fs, k)) {
       t->u.ind.idx = k->u.info;  /* literal string */
       t->k = VINDEXSTR;
@@ -1299,7 +1316,8 @@ static int validop (int op, TValue *v1, TValue *v2) {
     case LUA_OPBAND: case LUA_OPBOR: case LUA_OPBXOR:
     case LUA_OPSHL: case LUA_OPSHR: case LUA_OPBNOT: {  /* conversion errors */
       lua_Integer i;
-      return (tointegerns(v1, &i) && tointegerns(v2, &i));
+      return (luaV_tointegerns(v1, &i, LUA_FLOORN2I) &&
+              luaV_tointegerns(v2, &i, LUA_FLOORN2I));
     }
     case LUA_OPDIV: case LUA_OPIDIV: case LUA_OPMOD:  /* division by 0 */
       return (nvalue(v2) != 0);