| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Peter Eisentraut <peter(at)eisentraut(dot)org> |
| Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Update timezone to C99 |
| Date: | 2025-11-19 15:48:06 |
| Message-ID: | 1462381.1763567286@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Peter Eisentraut <peter(at)eisentraut(dot)org> writes:
> On 14.11.25 22:11, Tom Lane wrote:
>> "leapyear" is bool, and I believe these casts-to-int were put in to
>> suppress compiler bleats about up-casting that to int. This probably
>> dates from when we equated bool to char, and maybe it's moot now,
>> but I'm not sure.
> Correct, with bool as char this would have fallen afoul of
> -Wchar-subscripts, which is included in -Wall.
> I could remove these casts as a separate patch so that the reason can be
> documented more clearly.
Either that or just explain it in a para of the commit message.
>> Since we've dropped pre-C11 support, I wonder why we'd include this
>> upstream workaround for that.
> I figured it would be better to reduce divergences from upstream, even
> if we don't need all the code.
[ shrug... ] Okay.
No further comments.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Ilia Evdokimov | 2025-11-19 15:52:08 | Re: Use merge-based matching for MCVs in eqjoinsel |
| Previous Message | Tom Lane | 2025-11-19 15:43:58 | Re: PRI?64 vs Visual Studio (2022) |