"buffer too small" or "path too long"?

From: Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>
To: pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: "buffer too small" or "path too long"?
Date: 2022-06-13 03:05:51
Message-ID: 20220613.120551.729848632120189555.horikyota.ntt@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Recently we added the error messages "buffer for root directory too
small" and siblings to pg_upgrade. This means "<new_cluster's
pgdata>/pg_upgrade_output.d" was longer than MAXPGPATH.

I feel that the "root directory" is obscure here, and moreover "buffer
is too small" looks pointless since no user can do something on the
buffer length. At least I can't read out from the message concretely
what I should do next..

The root cause of the errors is that the user-provided directory path
of new cluster's root was too long. Anywhich one of the four buffers
is overflowed, it doesn't makes any difference for users and doesn't
offer any further detail to suppoerters/developers. I see "output
directory path of new cluster too long" clear enough.

Above all, this change reduces the number of messages that need
translation:)

# And the messages are missing trailing line breaks.

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center

Attachment Content-Type Size
0001-Revise-some-error-messages-of-pg_upgrade.patch text/x-patch 2.1 KB

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2022-06-13 03:24:42 Re: bogus: logical replication rows/cols combinations
Previous Message Masahiko Sawada 2022-06-13 02:58:57 Re: [BUG] Logical replication failure "ERROR: could not map filenode "base/13237/442428" to relation OID" with catalog modifying txns