Re: Patch proposal: New hooks in the connection path
| От | Zhihong Yu |
|---|---|
| Тема | Re: Patch proposal: New hooks in the connection path |
| Дата | |
| Msg-id | CALNJ-vSwQhS03xAYfftPxq_JnqY8KsMZzc9guCaB31Ge-jO3pQ@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: Patch proposal: New hooks in the connection path ("Drouvot, Bertrand" <bdrouvot@amazon.com>) |
| Ответы |
Re: Patch proposal: New hooks in the connection path
|
| Список | pgsql-hackers |
On Mon, Jul 4, 2022 at 5:54 AM Drouvot, Bertrand <bdrouvot@amazon.com> wrote:
Hi,
On 7/2/22 1:00 AM, Nathan Bossart wrote:
> Could we model this after fmgr_hook? The first argument in that hook
> indicates where it is being called from. This doesn't alleviate the need
> for several calls to the hook in the authentication logic, but extension
> authors would only need to define one hook.
I like the idea and indeed fmgr.h looks a good place to model it.
Attached a new patch version doing so.
Thanks
--
Bertrand Drouvot
Amazon Web Services: https://aws.amazon.com
Hi,
+ FCET_SPT, /* startup packet timeout */
+ FCET_BSP, /* bad startup packet */
+ FCET_BSP, /* bad startup packet */
Looking at existing enum type, such as FmgrHookEventType, the part after underscore is a word.
I think it would be good to follow existing practice and make the enums more readable.
Cheers
В списке pgsql-hackers по дате отправления: