| From: | Chao Li <li(dot)evan(dot)chao(at)gmail(dot)com> |
|---|---|
| To: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Cc: | Heikki Linnakangas <hlinnaka(at)iki(dot)fi> |
| Subject: | pg_resetwal: Fix handling of commit timestamp XIDs |
| Date: | 2026-09-29 06:09:40 |
| Message-ID: | F61E486B-746B-42F3-A532-B8D4F1519583@gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi,
While working on patch [1], I spotted two small issues in pg_resetwal.
1. The -c option accepts two XIDs. The first is parsed by strtouint32_strict(), but the second is parsed by strtoul(). This seems to be an oversight in 636c1914b48, which introduced strtouint32_strict().
2. The help message for the -c option says that “zero means no change”, but this behavior is missing on the current master branch. I checked the git history, and the behavior was lost in 978cf02bb8c, which added separate “given” flags for command-line options.
The fixes for both issues are tiny. See the attached patch for details.
Unfortunately, I reviewed both patches but didn't notice these issues at the time. Both commits were included in PG19, so this might be worth back-patching to 19.
[1] https://postgr.es/m/F7B47D8D-BFCB-464F-A737-DA796AE69D91@gmail.com
Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/
| Attachment | Content-Type | Size |
|---|---|---|
| v1-0001-pg_resetwal-Fix-handling-of-commit-timestamp-XIDs.patch | application/octet-stream | 2.0 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | shihao zhong | 2026-09-29 06:11:28 | Re: Parallel vacuum: wrong error context when the leader vacuums an index |
| Previous Message | Masahiko Sawada | 2026-09-29 06:08:37 | Re: Parallel vacuum: I/O timings in the log leave out the parallel workers |