| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Dean Rasheed <dean(dot)a(dot)rasheed(at)gmail(dot)com> |
| Cc: | Yugo NAGATA <nagata(at)sraoss(dot)co(dot)jp>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Dave Page <dpage(at)pgadmin(dot)org> |
| Subject: | Re: Numeric x^y for negative x |
| Date: | 2021-08-06 16:15:18 |
| Message-ID: | 2492411.1628266518@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Dean Rasheed <dean(dot)a(dot)rasheed(at)gmail(dot)com> writes:
> So the "test for overflow by reverse-conversion" obviously isn't
> working as intended, and it's going through power_var_int() when it
> shouldn't. I don't think there's anything wrong with that code, so I
> think this is a compiler bug.
Yeah, looks like one.
> I guess the best thing to do is just test the value against
> PG_INT32_MIN/MAX, which is what int84() does. There are 2 other places
> in numeric.c that use similar code to check for int16/32 overflow, so
> it's possible that they're broken in the same way on that platform,
> but they aren't covered by the regression tests, so it's also possible
> that they're OK. Anyway, something like the attached seems likely to
> be safer.
Looks plausible by eyeball (I've not tested).
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Japin Li | 2021-08-06 16:27:36 | Re: [BUG] wrong refresh when ALTER SUBSCRIPTION ADD/DROP PUBLICATION |
| Previous Message | Dean Rasheed | 2021-08-06 16:06:20 | Re: Numeric x^y for negative x |