Re: [HACKERS] WIP: Faster Expression Processing v4
От | Andres Freund |
---|---|
Тема | Re: [HACKERS] WIP: Faster Expression Processing v4 |
Дата | |
Msg-id | 20170323174004.a2ez7kygx5ngqxlp@alap3.anarazel.de обсуждение исходный текст |
Ответ на | Re: [HACKERS] WIP: Faster Expression Processing v4 (Tom Lane <tgl@sss.pgh.pa.us>) |
Ответы |
Re: [HACKERS] WIP: Faster Expression Processing v4
|
Список | pgsql-hackers |
Hi, On 2017-03-23 13:14:59 -0400, Tom Lane wrote: > Looking at some of the coding choices you made here, I see that you're > making a hard assumption that called functions never scribble on their > fcinfo->arg/argnull arrays. I do not believe that we've ever had such > an assumption before. I think we did that before, e.g. ExecEvalScalarArrayOp(). Think there's others too. > Are we comfortable with that? If so, we'd better document it. I think it's ok, but we indeed should document it. I recall a note somewhere... Can't find it anywhere however, might have misremembered a note about pass-by-ref arguments. fmgr/README? A note in FunctionCallInfoData's definition? Greetings, Andres Freund
В списке pgsql-hackers по дате отправления: