On 14.06.22 03:55, Michael Paquier wrote:
> On Tue, Jun 14, 2022 at 09:52:52AM +0900, Kyotaro Horiguchi wrote:
>> At Tue, 14 Jun 2022 09:48:26 +0900 (JST), Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com> wrote in
>>> Yeah, I feel so and it is what I wondered about recently when I saw
>>> some complete error messages. Is that because of the length of the
>>> subject?
>>
>> And I found that it is alrady done. Thanks!
>
> I have noticed this thread and 4e54d23 as a result this morning. If
> you want to spread this style more, wouldn't it be better to do that
> in all the places of pg_upgrade where we store paths to files? I can
> see six code paths with log_opts.basedir that could do the same, as of
> the attached. The hardcoded file names have various lengths, and some
> of them are quite long making the generated paths more exposed to
> being cut in the middle.
We have this problem of long file names being silently truncated all
over the source code. Instead of equipping each one of them with a
length check, why don't we get rid of the fixed-size buffers and
allocate dynamically, as in the attached patch.