| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Andrew Gierth <andrew(at)tao11(dot)riddles(dot)org(dot)uk> |
| Cc: | Marko Tiikkaja <marko(at)joh(dot)to>, pgsql-bugs(at)lists(dot)postgresql(dot)org, dejan(dot)petrovic(at)islonline(dot)com |
| Subject: | Re: BUG #15060: Row in table not found when using pg function in an expression |
| Date: | 2018-02-13 17:32:16 |
| Message-ID: | 24837.1518543136@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-bugs |
I wrote:
> Hm. So we could improve on the fix I proposed earlier if there were a
> way to tell GetCachedPlan "take a new snapshot even though there is an
> active snapshot". It's still expensive, but at least the code path where
> we use an existing generic plan doesn't get any added cost.
Here's a draft patch along those lines. Because it changes
GetCachedPlan's API, I'd be a bit worried about back-patching it; there
might be third-party code calling that directly. However, given that it
took so many years for anyone to notice a problem here, I think maybe
just fixing it in HEAD is fine.
regards, tom lane
| Attachment | Content-Type | Size |
|---|---|---|
| use-new-snapshot-for-planning-in-non-readonly-SPI.patch | text/x-diff | 19.3 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2018-02-13 18:58:00 | Re: 答复: response time is very long in PG9.5.5 using psql or jdbc |
| Previous Message | Tom Lane | 2018-02-13 14:53:59 | Re: BUG #15060: Row in table not found when using pg function in an expression |