| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Postgres hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Incorrect use of errcode_for_file_access in backend code |
| Date: | 2018-09-02 20:07:47 |
| Message-ID: | 20180902200747.GC1343@paquier.xyz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi all,
While looking at another patch for slot.c, I have noticed what looks
like incorrect use of errcode_for_file_access:
- errcode_for_file_access() is used with rmtree(), which makes no sense
as this comes from common/rmtree.c, and a warning already shows up using
%m.
- ERRCODE_DATA_CORRUPTED is not used to mention corrupted data, and
instead errcode_for_file_access() gets called.
Wouldn't something like the attached provide more adapted error
handling? That's mostly error handling beautification, so I would be
incline to just fix HEAD.
(I have noticed some inconsistent error string format in autoprewarm.c
on the way.)
Thoughts?
--
Michael
| Attachment | Content-Type | Size |
|---|---|---|
| slot-errors.patch | text/x-diff | 3.1 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Robert Haas | 2018-09-02 20:32:32 | Re: Stored procedures and out parameters |
| Previous Message | Michael Paquier | 2018-09-02 19:44:31 | Re: Bug in slot.c and are replication slots ever used at Window? |