在C聚合函数中使用PG_GETARG_POINTER是否安全？

Question

我正在为PostgreSQL编写一个C聚合函数，目前我正在使用PG_GETARG_POINTER，如下所示：

Datum my_transfunc(PG_FUNCTION_ARGS) {
       mystype *trans_data = (mystype *) PG_GETARG_POINTER(0);
       // Do some operations on trans_data pointer.
       // That make trans_data's memory content change. Then
       PG_RETURN_POINTER(trans_data);
}

我的代码有什么问题（内存，......）吗？

Answer 1

backend/executor/nodeAgg.c说：

We compute aggregate input expressions and run the transition functions
in a temporary econtext (aggstate->tmpcontext).  This is reset at least
once per input tuple, so when the transvalue datatype is
pass-by-reference, we have to be careful to copy it into a longer-lived
memory context, and free the prior value to avoid memory leakage.  We
store transvalues in another set of econtexts, aggstate->aggcontexts
(one per grouping set, see below), which are also used for the hashtable
structures in AGG_HASHED mode.  These econtexts are rescanned, not just
reset, at group boundaries so that aggregate transition functions can
register shutdown callbacks via AggRegisterCallback.

由于复制过渡值，因此必须为Datum（例如bytea），以便copyDatum可以正确复制。

由于您使用的是指针（可能在SQL中声明为internal），因此应该没问题; internal是按值传递的。

你必须确保你的mystype被分配在一个内存上下文中，该上下文在整个语句中存活但后来被清除。