Andres Freund <andres(at)anarazel(dot)de> writes:
> But are we actually prepared to assume that nothing ever leaks in
> receivers? That's, as far as I can tell, not a documented
> assumption.
There's a lot that's not documented there :-(
> We're also assuming that we don't leak into MessageContext over such
> cycles, which seems wrong. At the very least things like
> errdetail_params() are happy to leak into MessageContext.
This leak isn't in MessageContext; if it were, there likely wouldn't
have been a noticeable problem. It's leaking in the executor's
context over repeat ExecutorRun cycles in the same execution state.
We do have an expectation that there won't be per-row leakage in
the ExecutorState.
regards, tom lane