add global share string table

3rd/lua/lstring.c

@@ -148,10 +148,9 @@ void luaS_remove (lua_State *L, TString *ts) {
 /*
 ** checks whether short string exists and reuses it or creates a new one
 */
-static TString *internshrstr (lua_State *L, const char *str, size_t l) {
+static TString *queryshrstr (lua_State *L, const char *str, size_t l, unsigned int h) {
   TString *ts;
   global_State *g = G(L);
-  unsigned int h = luaS_hash(str, l, g->seed);
   TString **list = &g->strt.hash[lmod(h, g->strt.size)];
   for (ts = *list; ts != NULL; ts = ts->u.hnext) {
     if (l == ts->shrlen &&
@@ -162,6 +161,13 @@ static TString *internshrstr (lua_State *L, const char *str, size_t l) {
       return ts;
     }
   }
+  return NULL;
+}
+
+static TString *addshrstr (lua_State *L, const char *str, size_t l, unsigned int h) {
+  TString *ts;
+  global_State *g = G(L);
+  TString **list = &g->strt.hash[lmod(h, g->strt.size)];
   if (g->strt.nuse >= g->strt.size && g->strt.size <= MAX_INT/2) {
     luaS_resize(L, g->strt.size * 2);
     list = &g->strt.hash[lmod(h, g->strt.size)];  /* recompute with new size */
@@ -174,6 +180,7 @@ static TString *internshrstr (lua_State *L, const char *str, size_t l) {
   return ts;
 }
 
+static TString *internshrstr (lua_State *L, const char *str, size_t l);
 
 /*
 ** new string (with explicit length)
@@ -224,3 +231,216 @@ Udata *luaS_newudata (lua_State *L, size_t s) {
   return u;
 }
 
+/*
+ * global shared table
+ */
+
+#include "rwlock.h"
+#include "atomic.h"
+#include <stdlib.h>
+
+#define SHRSTR_SLOT 0x10000
+#define HASH_NODE(h) ((h) % SHRSTR_SLOT)
+
+struct shrmap_slot {
+	struct rwlock lock;
+	TString *str;
+};
+
+struct shrmap {
+	struct shrmap_slot h[SHRSTR_SLOT];
+	int n;
+};
+
+static struct shrmap *SSM = NULL;
+
+LUA_API void
+luaS_initshr() {
+	struct shrmap * s = malloc(sizeof(*s));
+	memset(s, 0, sizeof(*s));
+	int i;
+	for (i=0;i<SHRSTR_SLOT;i++) {
+		rwlock_init(&s->h[i].lock);
+	}
+	SSM = s;
+}
+
+LUA_API void
+luaS_exitshr() {
+	int i;
+	for (i=0;i<SHRSTR_SLOT;i++) {
+		TString *str = SSM->h[i].str;
+		while (str) {
+			TString * next = str->u.hnext;
+			free(str);
+			str = next;
+		}
+	}
+	free(SSM);
+}
+
+static TString *
+query_string(unsigned int h, const char *str, lu_byte l) {
+	struct shrmap_slot *s = &SSM->h[HASH_NODE(h)];
+	rwlock_rlock(&s->lock);
+	TString *ts = s->str;
+	while (ts) {
+		if (ts->hash == h && 
+			ts->shrlen == l &&
+			memcmp(str, ts+1, l) == 0) {
+			break;
+		}
+		ts = ts->u.hnext;
+	}
+	rwlock_runlock(&s->lock);
+	return ts;
+}
+
+static TString *
+query_ptr(TString *t) {
+	unsigned int h = t->hash;
+	struct shrmap_slot *s = &SSM->h[HASH_NODE(h)];
+	rwlock_rlock(&s->lock);
+	TString *ts = s->str;
+	while (ts) {
+		if (ts == t)
+			break;
+		ts = ts->u.hnext;
+	}
+	rwlock_runlock(&s->lock);
+	return ts;
+}
+
+static TString *
+new_string(unsigned int h, const char *str, lu_byte l) {
+	size_t sz = sizelstring(l);
+	TString *ts = malloc(sz);
+	memset(ts, 0, sz);
+	ts->tt = LUA_TSHRSTR;
+	ts->hash = h;
+	ts->shrlen = l;
+	memcpy(ts+1, str, l);
+	return ts;
+}
+
+static TString *
+add_string(unsigned int h, const char *str, lu_byte l) {
+	TString * tmp = new_string(h, str, l);
+	struct shrmap_slot *s = &SSM->h[HASH_NODE(h)];
+	rwlock_wlock(&s->lock);
+	TString *ts = s->str;
+	while (ts) {
+		if (ts->hash == h && 
+			ts->shrlen == l &&
+			memcmp(str, ts+1, l) == 0) {
+				break;
+		}
+		ts = ts->u.hnext;
+	}
+	if (ts == NULL) {
+		ts = tmp;
+		ts->u.hnext = s->str;
+		s->str = ts;
+		tmp = NULL;
+	}
+	rwlock_wunlock(&s->lock);
+	if (tmp) {
+		// string is create by other thread, so free tmp
+		free(tmp);
+	}
+	return ts;
+}
+
+static TString *
+internshrstr (lua_State *L, const char *str, size_t l) {
+  TString *ts;
+  global_State *g = G(L);
+  unsigned int h = luaS_hash(str, l, g->seed);
+  unsigned int h0;
+  // lookup global state of this L first
+  ts = queryshrstr (L, str, l, h);
+  if (ts)
+    return ts;
+  // lookup SSM again
+  h0 = luaS_hash(str, l, 0);
+  ts = query_string(h0, str, l);
+  if (ts)
+    return ts;
+  // If SSM->n greate than 0, add it to SSM
+  if (SSM->n > 0) {
+    ATOM_DEC(&SSM->n);
+    return add_string(h0, str, l);
+  }
+  // Else add it to global state (local)
+  return addshrstr (L, str, l, h);
+}
+
+LUA_API void
+luaS_expandshr(int n) {
+  ATOM_ADD(&SSM->n, n);
+}
+
+LUAI_FUNC TString *
+luaS_clonestring(lua_State *L, TString *ts) {
+  unsigned int h;
+  int l;
+  const char * str = getaddrstr(ts);
+  global_State *g = G(L);
+  TString *result;
+  if (ts->tt == LUA_TLNGSTR) 
+    return luaS_newlstr(L, str, ts->u.lnglen);
+  // look up global state of this L first
+  l = ts->shrlen;
+  h = luaS_hash(str, l, g->seed);
+  result = queryshrstr (L, str, l, h);
+  if (result)
+    return result;
+  // look up SSM by ptr
+  result = query_ptr(ts);
+  if (result)
+    return result;
+  // ts is not in SSM, so recalc hash, and add it to SSM
+  h = luaS_hash(str, l, 0);
+  return add_string(h, str, l);
+}
+
+struct slotinfo {
+	int len;
+	int size;
+};
+
+static void
+getslot(struct shrmap_slot *s, struct slotinfo *info) {
+	memset(info, 0, sizeof(*info));
+	rwlock_rlock(&s->lock);
+	TString *ts = s->str;
+	while (ts) {
+		++info->len;
+		info->size += ts->shrlen;
+		ts = ts->u.hnext;
+	}
+	rwlock_runlock(&s->lock);
+}
+
+LUA_API int
+luaS_shrinfo(lua_State *L) {
+	struct slotinfo total;
+	struct slotinfo tmp;
+	memset(&total, 0, sizeof(total));
+	int i;
+	int len = 0;
+	for (i=0;i<SHRSTR_SLOT;i++) {
+		struct shrmap_slot *s = &SSM->h[i];
+		getslot(s, &tmp);
+		len += tmp.len;
+		if (tmp.len > total.len) {
+			total.len = tmp.len;
+		}
+		total.size += tmp.size;
+	}
+	lua_pushinteger(L, len);
+	lua_pushinteger(L, total.size);
+	lua_pushinteger(L, total.len);
+	lua_pushinteger(L, SSM->n);
+	return 4;
+}