Re: Lua GC optimize of some tricks

[Xianfu Pan <pxfgod@...>]

I am not quietly familiar with `luaL_checktype`, Thank you for informing me.

It's that fields are not always identical. Some fields whether appear in table depend upon whether they exist in inputs.

> it seems cleaner to just set nil without the previous test. 

DO you mean `for (...) setnilvalue(gval(n));`

Yes, I have tried, like previous what `setnilvalue` does in array part, Just Set It Nil without test.But without previous test, it would crash! (Access violation writing location On Windows), Macro gval(n) returns a value like follow:

cell = 0x00007FF7E7321768 (this address denied to write!)

{

   value_ = {gc =0x0, p,f,i,n = 0x0},

   tt_ = 0

}

I suspect the manner traversing hash part goes wrongly. Otherwise, did i use `gval(n)` by mistake?

2017-01-11 20:31 GMT+08:00 Roberto Ierusalimschy <roberto@inf.puc-rio.br>:

> So we optimize our function; it looks like this
>
> `
> local tmp = {}
> function cfg_read_from_database()
> for k, v in pairs(tmp) do tmp[k] = v end
> ...
> return tmp
> end

I believe you mean "tmp[k] = nil" ?


> Here is the implementation of `table.reset`
> `
> static int treset(lua_State *L) {
> if (lua_istable(L, 1)) {
> lua_resettable(L, 1);
> }
> else {
> luaL_error(L, "bad argument to table.reset table expect got %s",
> lua_typename(L, lua_type(L, 1)));
> }
> return 0;
> }

static int treset (lua_State *L) {
  luaL_checktype(L, 1, LUA_TTABLE);
  lua_resettable(L, 1);
  return 0;
}


> LUA_API void lua_resettable(lua_State *L, int idx){
> StkId o;
> lua_lock(L);
> api_checknelems(L, 2);

*   Why 2? Probably you want 0 here (and therefore does not need the check)

> o = index2addr(L, idx);
> api_check(L, ttistable(o), "table expected");
> luaC_resettable(L, hvalue(o));
> lua_unlock(L);
> }


> void luaC_resettable(lua_State *L, Table *h) {
> /*need not to barrierback, because new value is NIL*/
> Node *n, *limit = gnodelast(h);
> unsigned int i;
> for (i = 0; i < h->sizearray; i++){
> TValue *cell = &h->array[i];
> setnilvalue(cell);
> }
> for (n = gnode(h, 0); n < limit; n++) {  /* traverse hash part */
> TValue *cell = gval(n);
> if (!ttisnil(cell))

*   it seems cleaner to just set nil without the previous test

> setnilvalue(cell);
> }
> }
> `
>
> Are there any issues/pitfalls of my implementation. Or any advices for me?
> Many thanks.

If the fields are all the same (the same "type" of table), you wouldn't
need to clean the table before setting them (the new values would
overwrite the old ones).

-- Roberto